Keynote Presentations

Teachers’ Perceptions of Text Features Associated with Generative AI Use in Student Writing

Jill V Jeffery (Leiden University)

Gustaf Skar (Norwegian University of Science and Technology)

Anne Holten Kvistad (Norwegian University of Science and Technology)

Steve Graham (Arizona State University)

The rapid release of generative AI tools has changed the way that teachers interact with student writing, raising the potential of new affordances but also challenges. Similar to the adoption of Internet resources decades ago, teachers must adapt quickly, as they are faced with the question, when assessing student writing, of how students are using new technologies for school writing tasks, and how their uses may support or detract from teachers’ goals for assigning writing in the first place. To investigate how teachers are responding to these issues, we surveyed 476 secondary school Norwegian teachers regarding their uses of AI to teach writing, their beliefs and ethical concerns about the use of AI for writing, and preparedness to use AI. We found that a majority of teachers indicated that their students used AI for multiple purposes, including for information gathering, and for enhancing grammar, syntax, and vocabulary in their writing. We also found that teachers slightly disagreed that AI had benefits for writing, moderately agreed they were concerned about ethical issues in using AI for writing, rated their preparation to teach writing using AI as minimal, and slightly disagreed that their students were prepared to use AI when writing (Graham et al., in press). To provide further insight into these quantitative findings, this paper presents results from a qualitative analysis of an open survey item in which teachers shared their perceptions of the text features that signal students’ use of generative AI in their writing, as well as their perceptions of how these features factor into their assessments of student writing quality.

Towards Historical Sign Language Linguistics 2.0

Victoria Nyst (Leiden University)

Sign languages differ importantly from spoken languages in their transmission. In communities with hereditary deafness, family-based transmission across generations predominates. Elsewhere, deaf schools typically serve as the main hubs for the transmission and diffusion of signs, and sign languages with a shared history of deaf education may share signs as a result. Moreover, a variety of school policies is found to have a large impact on sign language transmission. The complexity of sign language transmission and its impact on the development of a global web of related and unrelated sign languages requires a diversity of approaches to answer the question: How do sign languages change and spread over time, and how is this influenced by their transmission history? To answer this question, we need to combine a number of different research fields, including historical linguistics, historical sociolinguistic research, ethnographic fieldwork, and machine learning to scale up our observations. This includes the use of the many historical sign lists that were compiled in the first century of deaf education. Such lists are actually rediscovered in a growing number of countries, including for sign languages that are now extinct.

Much of this history of sign languages and their communities is stored in the memories of the oldest generation of deaf signers. Documenting these memories is urgent and vital for safeguarding the immaterial heritage of deaf communities.

This talk reviews how new insights in sign language and deaf studies, data science, and digital deaf studies can be leveraged to shed light on the histories of sign languages and their deaf communities.

Borderlands, boundaries and frontiers in historical sociolinguistics: Moving beyond the monolingual narrative 

Andreas Krogull (University of Zurich)

Recent research in the thriving field of historical sociolinguistics has increasingly acknowledged the fact that “[m]ultilingual societies were composed of multilingual individuals who used more than one language in their daily lives” (Pahta et al. 2018: 3). Investigating various aspects and phenomena of multilingualism and language contact, historical sociolinguists have contributed to alternative accounts of European language historiography (e.g. Franceschini et al. 2023; Pavlenko 2023). In the Early and Late Modern Low Countries, the Dutch-German borderlands constitute a particularly intriguing sociolinguistic space that challenges the traditionally monolingual narrative of ‘the’ history of Dutch (e.g. Krogull 2021).

A cultural and linguistic contact zone for many centuries, the borderland area in the northeastern Dutch provinces (e.g. Groningen, Drenthe) and the neighbouring parts ofnorthwestern Germany (e.g. East Frisia) offers a fertile ground for the exploration of multilingual (writing) practices in the past. This case study focuses on the long nineteenth century, which, in the European realm, is characterised by nation-building and border-makingprocesses, and the spread of standard language ideologies (e.g. Rutten 2019). When taking amore holistic, multilingual view ‘from below’, however, we can find ample evidence of actual language practices that do not seem to fit the monolingual ideal of this era.

Drawing on a selection of handwritten archival sources, primarily letters, and addressing concepts like borders, boundaries and frontiers, this lecture aims to investigate the use of multilingual repertoires, involving varieties of Dutch, High and Low German (and possibly more). It will be shown how individuals and families deployed their layeredlinguistic resources in hybrid and fluid ways in order to communicate, also across borders, and sometimes to the extent that the boundaries between distinct ‘named’ languages become blurry. More generally, the Dutch-German case raises questions about territorially demarcated, allegedly monolingual spaces as the default in language history writing.

Laboratory sociolinguistics: between variation and processing

Cesko Voeten (University of Amsterdam) 

Following a similar trend in psycholinguistics, multiple disciplines of linguistics are in the process of adopting more quantitative and hence experimental methods. Next to the well-known example of laboratory phonology, a currently upcoming paradigm in sociolinguistics is 'laboratory sociolinguistics' (Van de Velde et al 2021). My talk introduces the laboratory-sociolinguistic approach and positions it within its methodological context. I also discuss two case studies (one focused on individual differences and one focused on dialectometry) that demonstrate the potential of the approach. Finally, I sketch future directions for the field.