--- title: "amod: adjectival modifier — definition and key papers" description: "amod (adjectival modifier) is the Universal Dependencies relation that links a nominal head to an adjective or adjectival phrase that modifies its…" canonical: "https://research.elamur.ai/free/en/concepts/ud-amod/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/concepts/ud-amod/) [home](https://md.research.elamur.ai/free/en/) / [concepts](https://md.research.elamur.ai/free/en/concepts/) / amod: adjectival modifier # 🌐 amod: adjectival modifier Universal Dependencies · 7 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/) amod (adjectival modifier) is the Universal Dependencies relation that links a nominal head to an adjective or adjectival phrase that modifies its meaning, as in "red car". The adjective is the dependent and the noun or pronoun is the head. Also: adjectival modifier, amod, attributive adjective (traditional grammar) ## What it is In [Universal Dependencies](https://md.research.elamur.ai/free/en/concepts/universal-dependencies/), a sentence is a tree of typed relations between words. amod is the label for one common case: an adjective that narrows or describes a noun it attaches to. In "a cold morning", the word *morning* is the head and *cold* is its amod dependent. The theory behind UD builds its syntax from three kinds of units. Nominals name entities, clauses name events, and modifiers specify both. The relation label records what kind of phrase the dependent is, so the type of a nominal's modifier can be read off the label alone: amod for an adjectival one, acl for a clausal one, nmod for a nominal one ([Marneffe et al. 2021](https://md.research.elamur.ai/free/en/papers/marneffe-2021-universal-dependencies/)). The label goes back to the cross-linguistic typology of universal Stanford dependencies ([Marneffe et al. 2014](https://md.research.elamur.ai/free/en/papers/marneffe-2014-universal-stanford-dependencies-cross-linguistic/)). It was then carried into the UD v1 inventory of 40 relation types ([Nivre et al. 2016](https://md.research.elamur.ai/free/en/papers/nivre-2016-universal-dependencies-v1-multilingual-treebank/)) and kept in the v2 inventory of 37 universal types ([Nivre et al. 2020](https://md.research.elamur.ai/free/en/papers/nivre-2020-universal-dependencies-v2-evergrowing-multilingual/)). ### Neighbouring relations The boundaries matter more than the core case. In textbook terms, amod differs from several nearby relations: | Relation | Typical dependent | Contrast with amod | | --- | --- | --- | | acl | participle or clause with its own arguments | "the man *sitting* there" | | det / nummod | determiner / cardinal number | "*this* car", "*three* cars" | | nmod | nominal | "the roof *of* the house" | | compound | noun used as a modifier | "*steel* bridge" | A predicative adjective is not amod. In "the car is red", *red* is the head of the clause and *car* is its nsubj. Languages may add their own subtypes to universal relations ([Nivre et al. 2020](https://md.research.elamur.ai/free/en/papers/nivre-2020-universal-dependencies-v2-evergrowing-multilingual/)). Subtyped forms of the type amod:subtype are rare, however, and most treebanks use plain amod. ## Examples - "She bought a *new* phone." The relation is amod(phone, new). - "We fixed the *broken* window." A bare participle used attributively is still amod: amod(window, broken). - "Il porte une chemise *blanche*." In this French example the adjective follows the noun: amod(chemise, blanche). A pronoun can also be the head, as in "something *strange*", annotated amod(something, strange). An adjective can modify another adjective-headed nominal, as in "the *poor* old", where the noun is missing. ## Across languages [English](https://md.research.elamur.ai/free/en/topics/english/) places attributive adjectives before the noun. Its amod edges are overwhelmingly noun-to-adjective. Smaller groups have a proper noun as head ("*ancient* Rome"), a verb-tagged participle as dependent, a pronoun as head, or an adjective as head. [French](https://md.research.elamur.ai/free/en/topics/french/) shows the same dominance of noun heads with adjective dependents. Most of its adjectives follow the noun, and they agree with it in gender and number. A minority of edges have a numeral-tagged dependent, which reflects tagging choices for ordinal-like words. These cross-language differences are in word order and [morphology](https://md.research.elamur.ai/free/en/concepts/morphology/), not in the relation itself. That is the parallelism the scheme aims for ([Marneffe et al. 2021](https://md.research.elamur.ai/free/en/papers/marneffe-2021-universal-dependencies/)). Languages with a small or absent adjective class pose a harder question. There, property concepts are often expressed with verbs or relative clauses, and such modifiers fall under acl rather than amod. How UD treats predication that departs from the prototype is itself still being refined (Nivre et al. 2026). ## In treebanks and research UD treebanks store amod in the CoNLL-U format, together with lemma, [part-of-speech](https://md.research.elamur.ai/free/en/concepts/part-of-speech-tagging/) tag and features ([Nivre et al. 2016](https://md.research.elamur.ai/free/en/papers/nivre-2016-universal-dependencies-v1-multilingual-treebank/)). Because the label usually pairs a NOUN head with an ADJ dependent, its counts are a natural check on annotation. Inconsistency checks of UD corpora examine exactly such recurring head–dependent patterns ([Marneffe et al. 2017](https://md.research.elamur.ai/free/en/papers/marneffe-2017-assessing-annotation-consistency-universal-dependencies/)). For [dependency parsing](https://md.research.elamur.ai/free/en/concepts/dependency-parsing/), amod is one of the easier relations to predict. The adjective usually sits next to its head and is marked by a distinctive tag. The CoNLL shared tasks of 2017 and 2018 evaluated parsers on raw text in many UD languages, amod included ([Marneffe et al. 2021](https://md.research.elamur.ai/free/en/papers/marneffe-2021-universal-dependencies/)). Enhanced English UD keeps the basic tree in full and only adds edges or refines labels ([Schuster & Manning 2016](https://md.research.elamur.ai/free/en/papers/schuster-2016-enhanced-english-universal-dependencies-improved/)). As a result, amod edges carry over unchanged. Shared modifiers can also be spread to coordinated nouns, so that in "new cars and trucks" the adjective modifies both nouns. Outside parsing, amod defines a useful context slot for word meaning. One study of [language change](https://md.research.elamur.ai/free/en/topics/language-change/) parses diachronic corpora with Stanza and compares the adjectival-modifier fillers of a target word across periods using Jensen-Shannon divergence ([Tat et al. 2026](https://md.research.elamur.ai/free/en/papers/tat-2026-transparent-semantic-change-detection-dependency/)). For English *plane*, modifiers such as *military* rose and *horizontal* declined. This traces the spread of the aircraft sense, and each such detected shift can be checked against the specific collocates that drove it ([Tat et al. 2026](https://md.research.elamur.ai/free/en/papers/tat-2026-transparent-semantic-change-detection-dependency/)). How often the relation amod 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 | 3.93% of tokens (44,237 tokens) | | French treebanks | 3.67% of tokens (40,050 tokens) | What amod joins: the most frequent head tag → dependent tag The five most frequent in each language's treebanks, by tokens. **The numbers** | Pair | Tokens | | --- | --- | | English: NOUN → ADJ | 37,846 | | English: PROPN → ADJ | 2,548 | | English: NOUN → VERB | 2,005 | | English: PRON → ADJ | 507 | | English: ADJ → ADJ | 298 | | French: NOUN → ADJ | 37,356 | | French: PROPN → ADJ | 1,304 | | French: NOUN → NUM | 403 | | French: PRON → ADJ | 374 | | French: NUM → ADJ | 99 | ## 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 - [Transparent Semantic Change Detection with Dependency-Based Profiles](https://md.research.elamur.ai/free/en/papers/tat-2026-transparent-semantic-change-detection-dependency/) — Bach Phan Tat, Kris Heylen, Dirk Geeraerts et al. · 2026 · 1 citation ## 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 advmod: adverbial modifier](https://md.research.elamur.ai/free/en/concepts/ud-advmod/) - [next concept → appos: appositional modifier](https://md.research.elamur.ai/free/en/concepts/ud-appos/) - [↑ back to concepts](https://md.research.elamur.ai/free/en/concepts/)