Is the framework of Universal Dependencies (UD) compatible with findings from linguistic typology? To address this question, we need to systematically review how UD represents linguistic constructions in the world’s languages, and how it handles the range of morphosyntactic variation attested in linguistic typology. In this paper, we start this review by discussing reference and modification constructions. The review shows that, although UD can represent all major constructions in this area, there are a number of cases where UD categories do not align systematically with a typological classification of constructions, and where constructional similarity is therefore not transparent across languages. We also identify limitations in the representation of certain morphosyntactic strategies, notably indexation and linkers. To overcome these limitations, we propose a number of revisions that may be considered for future versions of UD.

Relative forms are a distinctive morphosyntactic feature of Earlier Egyptian. They pose a challenge when annotating them according to the Universal Dependencies approach. They are adjective finite verb forms, and therefore they have both verb and adjective properties, but they can also be used as nouns. The aim of this paper is to discuss the morphosyntactic methodology applied to their annotation in the Egyptian-UJaen treebank.

This paper describes the details of UD-based morphological and syntactic annotations on Esperanto texts to construct its small-scale UD treebank. Though it was created as an international auxiliary language, Esperanto has increasingly been studied as a natural language both in linguistics and in NLP. This paper introduces the detail of manual annotation of UD morphological and relational tags and describes how the frequencies of these tags differ across the treebanks and discusses the possibility of future research of Esperanto as a natural language.

This contribution presents the Naga-Suansu Universal Dependencies (UD) treebank, the first resource of this kind for Suansu, an endangered and underdocumented Tibeto-Burman language spoken in Northeast India. This treebank follows the UD annotation framework. We describe the corpus composition, data sources, and annotation process, outlining the general structure of the treebank. In addition, we highlight morphosyntactic challenges where Suansu grammar does not fit neatly into the UD annotation schema and propose adaptations to better capture its structural properties. As the first Tibeto-Burman language included in the UD project, the Naga-Suansu treebank serves several purposes: it contributes to the documentation and preservation of endangered languages, enables the understanding of cross-linguistic variation, and supports future research efforts in refining UD annotation practices for South and Southeast Asian languages.

This paper presents the first treebank for the dialect of Lesbos, a low-resource living Northern variety of Modern Greek (MG), annotated according to the Universal Dependencies (UD) framework. So far, the only dialectal treebank available for Greek developed with cross-dialectal knowledge transfer is an East Cretan one, which belongs to the same Southern branch as Standard Modern Greek (SMG). Our study investigates the effectiveness of cross-dialectal knowledge transfer between dialectologically less similar varieties of the same language by leveraging knowledge from SMG to annotate the Northern dialect of Lesbos. We describe the annotation process, present the resulting treebank, inject additional linguistic knowledge to enhance the results, and evaluate the effectiveness of cross-dialectal knowledge transfer for active annotation. Our findings contribute to a better understanding of how dialectal variation within language families affects knowledge transfer in the UD framework, with implications for other low-resource varieties.

CHILDES is a widely used resource of transcribed child and child-directed speech. This paper introduces UD-English-CHILDES, the first officially released Universal Dependencies (UD) treebank. It is derived from previously dependency-annotated CHILDES data, which we harmonize to follow unified annotation principles. The gold-standard trees encompass utterances sampled from 11 children and their caregivers, totaling over 48K sentences (236K tokens). We validate these gold-standard annotations under the UD v2 framework and provide an additional 1M silver-standard sentences, offering a consistent resource for computational and linguistic research.

Despite recent advances in digital resources for other Coptic dialects, especially Sahidic, Bohairic Coptic, the main Coptic dialect for pre-Mamluk, late Byzantine Egypt, and the contemporary language of the Coptic Church, remains critically under-resourced. This paper presents and evaluates the first syntactically annotated corpus of Bohairic Coptic, sampling data from a range of works, including Biblical text, saints’ lives and Christian ascetic writing. We also explore some of the main differences we observe compared to the existing UD treebank of Sahidic Coptic, the classical dialect of the language, and conduct joint and cross-dialect parsing experiments, revealing the unique nature of Bohairic as a related, but distinct variety from the more often studied Sahidic.

In this paper we study the representation of negation in UD treebanks. We show that the existing annotations are often inconsistent with the guidelines and that there are ill-motivated differences in annotation of constructions across and even within languages. Moreover, we argue that even if the annotation of the two negation-related features (Polarity=Neg and PronType=Neg) were consistent, these two features would be inadequate for straightforwardly expressing the semantics of negation because they relate to the word level only and hence to form rather than meaning. We therefore propose to add two features, Negated=+ and DoubleNegated=+, which directly encode when a predicate is semantically under negation, and thereby allow a straightforward semantic interpretation of a UD parse in terms of negation.

The increasing complexity of environmental discourse is directly proportional to the growing complexity of environmental debates present today in all communication media. While linguistic and communication studies have been pursued on this discourse, the development of computational linguistic tools and resources dedicated to support its analysis and interpretation is still very incipient. For one, no morphosyntactic resources specific to the environmental domain can be found on major platforms and repositories. This paper introduces TreEn, a multilingual treebank project in progress which compiles texts on environmental discourse produced in different conversational and communication contexts. In particular, it reports on the parallel component of the project and discusses issues faced during sentence-level alignment between original and translated texts, annotation of texts following UD guidelines, and labeling entities drawing on an ontology of environmental-related topics. This novel resource is expected to support environmental discourse analysis by providing morphological and syntactical data to enable cross-language and cross-cultural comparison based on the semantics of the entities annotated in the treebank.

In this paper we present a sample treebank for Old English based on the UD Cairo sentences, collected and annotated as part of a classroom curriculum in Historical Linguistics. To collect the data, a sample of 20 sentences illustrating a range of syntactic constructions in the world’s languages, we employ a combination of LLM prompting and searches in authentic Old English data. For annotation we assigned sentences to multiple students with limited prior exposure to UD, whose annotations we compare and adjudicate. Our results suggest that while current LLM outputs in Old English do not reflect authentic syntax, this can be mitigated by post-editing, and that although beginner annotators do not possess enough background to complete the task perfectly, taken together they can produce good results and learn from the experience. We also conduct preliminary parsing experiments using Modern English training data, and find that although performance on Old English is poor, parsing on annotated features (lemma, hyperlemma, gloss) leads to improved performance.

Sindhi is an Indo-Aryan language spoken primarily in Pakistan and India by about 40 million people. Despite this extensive use, it is a low resource language for NLP tasks, with few datasets or pretrained embeddings available. In this work, we explore linguistic challenges for annotating Sindhi in the UD paradigm, such as language-specific analysis of adpositions and verb forms. We use this analysis to present a newly annotated dependency treebank for Universal Dependencies, along with pretrained embeddings and an annotation pipeline specifically for Sindhi annotation.

This paper reports on the development of the first dependency treebank for Khoekhoe (KDT). Khoekhoe (Khoe-Kwadi, Namibia) is a low-resource language with few linguistic and computational resources available publicly. This treebank consists of 29k words across six texts taken from various registers. It includes a substantial portion of spoken conversational data. These sentences were annotated manually according to the Universal Dependencies framework. In this paper, apart from presenting the strategies that have been followed to create the treebank, we also discussed some challenging morphological features and syntactic constructions found in the corpus and outlined how we have handled them using the current Universal Dependencies specification.

We introduce the first fully aligned and manually annotated parallel Universal Dependencies (UD) treebanks for four Turkic languages: Azerbaijani, Kyrgyz, Turkish, and Uzbek. These resources currently consist of 148 strategically selected sentences that illustrate typologically significant morphosyntactic phenomena across these related yet distinct languages. These parallel treebanks enable systematic comparative studies of Turkic syntax and may be instrumental in cross-lingual NLP applications. All treebanks are available as part of UD v2.16.

Austronesian languages exhibit features that are challenging for Universal Dependencies: most notably, the symmetric voice system, whereby agent, patient, and instrumental arguments (among others) can be the pivot of a transitive structure – complicating the usual assumption that subjects of transitive sentences are semantic agents, and objects semantic patients. To showcase our ideas of how to address the representation of such systems in Universal Dependencies, we introduce a small treebank of sentences from texts and elicitation sessions in Gorontalo, an Austronesian language of Sulawesi (Indonesia), which exhibits a Philippine-type voice system. We discuss the annotation guidelines for this language, and the extensions of the Universal Dependencies guidelines that are needed to accommodate this and other Austronesian languages.

Enhanced Universal Dependencies (EUD) serve as a crucial link between syntax and semantics. Beyond basic syntactic dependencies, EUD provides valuable refined logical connections for downstream tasks such as semantic role labeling, coreference resolution, information extraction, and question answering. The original EUD framework defines six types of relationships, but this paper introduces an extension designed to address subject propagation in pro-drop languages. This “Extended EUD” proposal increases the number of relationships that may be annotated in sentences, improving linguistic representation. Additionally, we report our experiments on a corpus of Portuguese (a pro-drop language), which we make publicly available to the research community.

Universal Dependencies (UD) is gaining popularity as an annotation standard for second language (L2) material. Grammatical errors and other interlanguage phenomena, however, pose significant challenges that official guidelines only address in part. In this paper, we give an overview of current annotation practices and provide some suggestions for harmonizing guidelines for learner corpora.

This paper presents a Universal Dependencies (UD) treebank of Gwich’in, a severely endangered Athabascan language. The treebank, developed using instructional materials and dictionaries, includes 313 annotated sentences. This paper discusses the methodology used to construct the treebank, the linguistic challenges faced, and the implications of annotating a polysynthetic, morphologically complex language within the Universal Dependencies framework. The treebank was released with UD version 2.15 and available at https://github.com/UniversalDependencies/UD_Gwichin-TueCL/.

Words, more specifically “syntactic words”, are at the centre of a dependency-based approach like Universal Dependencies. Nonetheless, its guidelines do not make explicit how such a word should be defined and identified, and so it happens that different treebanks use different standards to this end. To counter this vagueness, the community has been recently discussing a definition put forward in (Haspelmath, 2023) which is not fully uncontroversial. This contribution is a preliminary case study that tries its hand at concretely applying this definition (except for compounds) to Latin in order to gain more insights about its operability and groundedness. This is helped by the spread of Latin over many treebanks, the presence of good linguistic resources to analyse it, and a linguistic type which is probably not fully considered in (Haspelmath, 2023). On the side, this work shows once more the difficulties of turning theoretical definitions into working directives in the realm of linguistic annotation.

This paper introduces ShUD, the first Universal Dependencies (UD) treebank for Shanghainese, a Wu Chinese variant spoken by approximately 14 million people but severely under-resourced in NLP. The treebank is built through a scalable annotation pipeline that exploits grammatical parallels between Shanghainese and Mandarin. Our pipeline also provides a practical strategy for bootstrapping resources for other Chinese dialects. We documented syntactic phenomena unique to Shanghainese within the UD framework and fine-tuned a dependency parser using our annotated treebank, contributing a foundation to both NLP tool development and cross-linguistic syntactic research.