What do we do?
MEDAL activities are grouped into training, research and outreach.
Training and resources for early-career researchers
Summer Schools
During the project, there will be three summer schools, each corresponding to a particular methodology.
- Year 1: Corpus Linguistics
- Year 2: Experimental Methods
- Year 3: Computational Modelling
Training Events
In addition to the annual summer schools, we organise workshops and master classes on topics such as:
- Empirical methods
- Open Science
- Soft skills for career building
- Statistics
Final conference
In the final year of the MEDAL project, we will organise a final conference in Tartu to showcase the methodological diversity of linguistics today.
In order to make these events as accessible as possible, most events are either fully online, or as hybrid as is possible in any given situation. Lectures are also recorded. All materials such as slides, handouts or codes that are used by instructors during workshops can be found under the ‘Resources’ section of any given event on the Events page. In addition, we have a Resources page for useful external resources, such as online language databases.
You can visit our Events page to see our upcoming events.
You can visit our Resources page to find useful external resources.
Collaborative research: GoLD MEDAL
Besides providing methodological training, MEDAL also offers room to put the learned skills into practice: Part of the project encompasses a sub-project dedicated entirely to research, entitled Generalising over Linguistic Data (GoLD). This part of the project is referred to as GoLD MEDAL.
There are currently three different studies being conducted within this project. Reflecting the three main methodologies of MEDAL (Corpus, Experiments & Computational Modeling), the first phase of the study involves a parallel investigation of corpus data from multiple languages, in line with the first year’s training school in corpus methodology, including cross-linguistic and multimodal corpus data available to network members, to generate hypotheses based on disparate language data. Subsequently, the research team will model the data and design an experiment to test the model across the languages aligned with the training school in experimental methods. The final research phase, in step with the final training school, will involve testing computational models against experimental results, and feeding this back, full circle, into the corpus data.
The three primary purposes of the research project are:
- To offer early-career researchers and other research staff an opportunity to apply their newly acquired skills and knowledge.
- To establish research collaboration between all four partners, leading to joint academic output and visibility for all partners.
- To derive high-quality, large-scale research results on a broad sample of languages, thereby contributing to a more representative and inclusive research landscape in Europe. The research will be cross-linguistic (at least two language families and three modalities – written, spoken and signed), multimethod, and interdisciplinary.
Currently, there are three ongoing GoLD research projects amongst MEDAL consortium members.
The three ongoing GoLD research topics
This research uses data from Communicative Development Inventories (CDIs) data from Estonian and other morphologically rich languages to address a long-standing question in language development: What the relationship between early lexical and grammatical knowledge? While there is a strong relationship between vocabulary and syntax in early language development, the two are separable from about 18 months. We are investigating whether the same patterns hold for morphology when we compare cross-linguistic data.
This study involves Virve Vihman, Adele Vaks, Piia Taremaa and Ada Urm from the University of Tartu; Caroline Rowland, Anastasia Chuprina and Izabela Jordanoska from the Max Planck Institute for Psycholinguistics and Radboud University in Nijmegen; and Seamus Donnelly from the Australian National University.
It has long been known that, in languages that allow subject omission, the subject is not dropped in all cases. What factors in a language can account for pro-drop? In this corpus-based study, we compare pro-drop of the first and second person in conversational data of Finno-Ugric and Slavic languages.
This study involves Mari Aigro, Maarja-Liisa Pilvik, Piia Taremaa, Virve Vihman and Joshua Wilbur from the University of Tartu; Anastasia Chuprina and Izabela Jordanoska from the Max Planck Institute for Psycholinguistics and Radboud University in Nijmegen; and Dagmar Divjak, Petar Milin and Justyna Mackiewicz from the University of Birmingham.
This study examines expressions of negation using a multimodal conversational corpus of Estonian annotated for both speech and gesture, and comparing that to Turkish. We look at which gestures occur with spoken negators and when and how negation is expressed in the visual modality with manual and non-manual (head, eyebrows, etc.) markers. We examine how this relates to both syntax and information structure. We also make use of automatic kinematic analyses using OpenPose (Cao et al. 2017) and ELAN (Sloetjes & Wittenburg 2008).
This study involves Pärtel Lippus, Eva Liina Asu-Garcia, Maarja-Liisa Pilvik and Liina Lindström from the University of Tartu; and Aslı Özyürek, Sevilay Şengül Yıldız and Hatice Zora from the Max Planck Institute for Psycholinguistics and Radboud University in Nijmegen.uistics.
Outreach events & science communication
Another MEDAL aim is to promote linguistics outside academia, particularly through activities aimed at secondary school students and general audiences in each of the consortium’s countries.
Parallel to each summer school, we organise a masterclass for secondary school students on the topic of the respective summer school. The topic of the 2023 Summer School was understanding the power of language by using freely available online corpora to find common collocations through hands-on case studies on terms that play a key role in current media discourse such as “(im)migrant” versus “refugee”.
The 2024 Summer School will feature a masterclass on psycholinguistics and the 2025 one on computational linguistics.
In addition, for the final year, we will organise a popular science festival in Nijmegen! Stay tuned!
Under our Resources page, you can also find some popular science outlets and blogs that feature work of MEDAL consortium members.