Constructicography in Practice: Constructions, Challenges, and Opportunities
Co-located with EURALEX 2026 - 29.09.2026 - Vienna, Austria
Constructicography (Lyngfelt et al. 2018), the systematic description of constructions in Constructicons and construction-based lexicographic resources, has emerged as a key area at the intersection of Construction Grammar (i.e., Croft 2001), lexicography, and computational linguistics. While constructions have long been acknowledged as central units of linguistic knowledge, their consistent representation, annotation, and integration into lexicographic resources remain methodologically and technologically challenging.
The workshop aims to bring together researchers working on theoretical, descriptive, and computational aspects of constructicography. It will focus on how constructions are identified, modeled, represented, and linked to lexical, semantic, and pragmatic information across languages and resource types. Particular attention will be paid to challenges such as construction granularity, variation, productivity, cross-linguistic comparability, corpus-driven discovery, and interoperability with existing lexical resources.
At the same time, the workshop highlights emerging opportunities offered by large corpora, annotation frameworks, and NLP techniques, including construction mining, semi-automatic Constructicon building, and the use of large language models in constructional analysis. Moreover, the workshop will explore applications of Constructicons, particularly in language teaching and learning.
By fostering dialogue among linguists, lexicographers, educational experts, and computational researchers, the workshop seeks to advance constructicography as a mature, practically applicable field within lexicography.
At the same time, the workshop highlights emerging opportunities enabled by large corpora, annotation frameworks, and NLP techniques, including construction mining, semi-automatic Constructicon development, and the use of large language models in constructional analysis. In addition, the workshop will explore applications of Constructicons, particularly in language teaching and learning.
We invite submissions on (but not limited to) the following topics:
- Theoretical foundations of constructicography
- Methods for Constructicon (and FrameNet) development
- Frame-based approaches to constructional meanings
- Corpus-based identification and annotation of constructions (and, possibly frames)
- Defining and delimiting constructions
- (Limits of) capturing constructional meanings by frames
- Comparing frame-based to other approaches to constructicon building
- Construction annotation in treebanks
- Construction annotation in treebanks - universality of constructions
- Constructions and Multiword Expressions
- Constructions and frames
- Computational approaches to constructicography
- Constructicons, constructions, frames, and LLMs
- Applications in language learning, teaching, downstream NLP applications, and beyond
Authors of accepted abstracts will be invited to submit full papers for inclusion in a dedicated volume after the workshop.
Invited Speaker
We are pleased to announce that Jelena Kallas (PHD) will give an invited talk about Constructions, Collo-Profiles, and CEFR Levels: A Hybrid Corpus-Based and AI-Assisted Workflow for Constructicon Building.
Jelena Kallas (PhD) is a senior researcher at the Institute of the Estonian Language. Her research focuses on corpus lexicography, pedagogical lexicography, constructicography, and L2 teaching and learning applications. She has been involved in the development of major modern Estonian lexicographic resources and corpora, including the EKI Combined Dictionary, the Estonian Vocabulary and Grammar Profile, and the Estonian National Corpus.
She currently leads the research project "Expanding the Scope of a Multi-Purpose Lexicographic Resource to Grammar and L2 Competence", which aims to develop methodological foundations for representing grammar and linguistic competence in an integrated multi-purpose lexicographic resource. The project also includes the development of the Estonian Constructicon. Her recent work focuses particularly on corpus-based and AI-assisted methods for identifying and describing constructions, their collo-profiles, and CEFR-related information.
Jelena has extensive experience in European lexicographic collaboration. She was a member of the EURALEX Executive Committee from 2014 to 2021 and a member of the organising and scientific committees of the eLex conferences from 2013 to 2023. She represents Estonia in several international research networks and infrastructures, including the ELEXIS Association, CLARIN-ELEXIS, UniDive, and ENEOLI. She is also involved in ELEXAI (European Lexicographic Infrastructure for Artificial Intelligence), a Horizon Europe project developing an AI-enhanced European lexicographic infrastructure and exploring how high-quality lexicographic knowledge can contribute to large language models.
Preliminary programme
| 9:00 – 9:10 | Opening and welcome Voula Giouli, Athina Sioupi, Nina Böbel, Alexander Ziem |
| 9:10 – 10:00 | Invited talk: Constructions, Collo-Profiles, and CEFR Levels: A Hybrid Corpus-Based and AI-Assisted Workflow for Constructicon Building Jelena Kallas |
| 10:00 – 10:20 | The CASA-ConstructiCon of the English language – the project at present and plans for the future Thomas Herbst |
| 10:20 – 10:40 | Building a Constructicon for an Ancient Language: Challenges and Prospects from ConNetGk Chiara Zanchi |
| 10:40 – 11:00 | Elodie Winckel, Peter Uhrig, Stephanie Evert |
| 11:00 – 11:30 | COFFEE & TEA BREAK |
| 11:30 – 11:50 | Elizaveta Petrenko, Elodie Winckel, Peter Uhrig |
| 11:50 – 12:10 | Francesca Masini, Ludovica Pannitto, Francesco Massidda, Flavio Pisciotta, Elisa Chierchiello, Cristina Bosco, Bruno Guillaume |
| 12:10 – 12:30 | Silvia Luraghi |
| 12:30 – 12:50 | Illustrating filler-slot relations Bálint Sass |
| 12:50 – 13:00 | BREAK |
| 13:00 – 13:20 | Paul Sambre |
| 13:20 – 13:40 | Anna Endresen, Olaf Mikkelsen, Chadi Ben Joussef |
| 13:40 – 14:00 | Eleni Kanli, Evelyn Wiesinger |