20 & 21 August 2025
Structure and Activities
Camp 3 was structured across three days of intensive lectures, workshops, and collaborative discussions. The camp’s opening session provided the project’s Research Leads and Supervisors with an opportunity to update the wider team on the latest research developments at each institution. Further sessions combined theoretical perspectives with practical exercises, allowing ESRs to critically engage with both methodological design and research application. The camp also provided less formal opportunities for ESRs to discuss their individual projects and receive feedback from peers, supervisors, and invited scholars. In keeping with CASCADE’s ethos, social events, including shared meals and a group boat trip reinforced community building and informal knowledge exchange. During this trip, the ESRs benefited from small-group discussions with post-doctoral researchers who were in a position to offer advice and insights into the life of a researcher beyond the PhD stage.
Training Camp 3 Programme Format
This training camp consisted of three different types of activities:
- Introductions and Training sessions where selected members and partners will provide
input to topics related to research questions and methods, followed by discussion. - Hands-on practicals, whereby ESRs worked on computational challenges designed by
fellow ESRs and senior researchers in their field. - A joint seminar with the University of Helsinki Summer School of Conceptual History
featuring presentations by two CASCADE project supervisors and two senior researchers
in Computational History.
The purpose of this training camp is to:
● Build on the ESRs understanding of the interplay between data, questions and methods with participation of research colleagues from King’s College London (Mark
J.Hill), the University of Turku (Antti Kanner) and the University of Helsinki Computational History Group
● Facilitate a reflection of experiences between the CASCADE ESRs and digital history doctoral students associated with the University of Helsinki
● Exchange more information about each other’s research
● Ask questions (about anything)
● And enjoy!
CAMP 3 Questions & Methods – Schedule of Events – 20 August 2025
09.15 –10:30 Prof. Mikko Tolonen – Opening Session
● What is CASCADE and a Marie Curie Training Network?
● Introducing the Helsinki Camp (people and the programme + working methods)
● Update from each institution on the (PhD) projects (PIs 5 min)
10.30 – 13.00 Session 1: Prof. Eetu Mäkelä – Using data to answer questions
10:30 – 11:15 What to take into account when answering research questions based on data?
11:15 – 11:30 Break
11:30 – 12:30 Hands-on: Digging into problems and confounders in data (with the Royal Society Corpus)
12:30 – 13:00 Common Debriefing
13:00 – 14:00 Lunch break
14:00 -17.00: Session 2: Antti Kanner and Mark Hill – Words and Concepts
Contrasting different computational approaches to the analysis of text and the types of relationships and patterns they uncover.
● What do linguistic features tell us about meaning at the surface level (for non-linguistics)?
● The importance of metadata: the more of it you have, the simpler the analysis
● Operationalising research questions about language / different linguistic methods depending on data
● An example of these things in practice: Case Puclic(k)
14:45 – 15:00 Break
15:00 – 16:30 Hands-on: Analysing the language within the Royal Society Corpus with R
16:30 – 17:00 Common Debriefing
19:00 Dinner
CAMP 3 Questions & Methods – Schedule of Events – 21 August 2025
9.00 – 13.00 Session 3: Prof. Eetu Mäkelä, Lidia Pivovarova, Ke Shu & Yu Wu
The unreasonable effectiveness of language models for textual analysis
● How to use LLMs for textual analysis in practice
● Don’t trust the AI: the research integrity -threatening pitfalls of relying on AI for textual analysis and how to avoid them
09:00 – 09:20 Ke Shu & Yu Wu (University of Helsinki)
Talk: Detecting Latin snippets in texts with Large Language Models
09:20 – 09:45 Lidia Pivovarova (University of Helsinki)
Talk: Morphological analysis and translation of Finnic oral poetry with LLMs
09:45 – 10:45 Hands on: Named Entity Recognition (NER) task using open-source Large Language Models (LLM) with historical text (CSC project)
10:45 – 11:00 Common Debriefing 1
11:15 – 12:30 Prof. Eetu Mäkelä: Hands-on: Avoiding LLM pitfalls
12:30 – 13:00 Common Debriefing 2
13:00 – 14:00 Lunch break
14:00 – 18:00 Seminar: Semantic Drift and Intellectual History
(Joint Seminar with the Summer School of Conceptual History)
Chair: Prof. Mikko Tolonen
● ‘The Concert of Disciplines, or, A Discursion on the Symphonious Study of Words and History’ | Prof. Dirk Geeraerts (KU Leuven)
● ‘What’s linguistics got to do with it? Reflections on interdisciplinary approaches to the history of ideas.’ | Prof. Susan Fitzmaurice (University of Sheffield)
● ‘From Structure to Semantics: Using Networks to Detect Linguistic Differences within Historical Communities’ | Mark Hill (King’s College London)
● ‘What are concepts in conceptual history? Revisiting Koselleck through theories of semantic relations’ | Jani Marjanen ( University of Helsinki ) & Antti Kanner (University of Turku)