Cognitive ergonomics of discourse
How textual structure, coherence cues, and discourse-level organization change the cost profile of interpretation for human readers.
About
I work at the intersection of cognitive science, computational linguistics/NLP, human–machine interaction, and philosophy of mind. My central interest is the cognitive and physiological “threshold moments” where structure emerges: critical points between chaos and order, randomness and recognition, noise and meaning.
Alongside academic work, I’ve led and built applied data-science and AI initiatives in higher education and industry (including program design, organizational transformation, and research advisory). I also maintain an enduring interest in craft and skill: knitting, mental model building, and the patient work of learning systems—human and technical—well enough to improve them.
Now
- M.A. student in Linguistics, University of Oslo (ongoing)
- Executive in Residence (NLP & Data Analysis), BI Norwegian Business School (ongoing)
- Teaching Assistant (Statistics in R for language students), University of Oslo (ongoing)
- Co-founder / daglig leder / researcher, Kanna AS (ongoing)
Research keywords: Discourse processing · meaning & context update · cognitive ergonomics · interpretive load · generative AI · human–machine interaction · philosophy of mind · experimental design · text analysis
Research
Research focus
My work asks what happens to human understanding when fluent language is cheap. In contemporary cognitive ecologies, people can share public symbols with AI systems while lacking shared grounding and stable interpretive mappings. This creates a practical and theoretical problem: coordination can look successful while the interpretive work, and its costs, shift in non-obvious ways.
I approach these questions from
- Cognitive science, including representational and predictive frameworks for how agents stabilize meaning under uncertainty.
- Computational linguistics/NLP, focusing on discourse, context management, and measurable properties of text that track interpretive effort.
- Philosophy of mind and language, focusing on what it means to understand, to be responsible for commitments, and to coordinate meaning with others.
Current themes
Meaning under AI mediation
How generative text affects attention, memory encoding, and epistemic agency—especially when users treat fluent output as a substitute for grounded understanding.
Threshold dynamics: noise → structure
Where and how interpretation “tips” from confusion into recognition (and back), and what linguistic cues reliably move the system across that boundary.
Methods and competencies
- Research design in behavioral/cognitive experiments: task design, manipulation design, operationalization, construct validity
- Mixed-methods analysis: quantitative modeling + structured qualitative analysis
- Text analysis in Python (incl. NLP) and statistics/visualization in R
Publications & Writing
Peer-reviewed / academic and evaluated work
- Rasmussen, J., Knutsen (Karajanov), J., & Arnulf, J. K. (2024). Styrer og bærekraft: Norske børsnoterte selskap møter forventninger med kontroll heller enn strategi. MAGMA. (Contributed code and classification logic for governance-theory-driven language grouping.)
- The Geography of Doubt: Demarcation, Anthropocentric Vulnerability, and the Epistemic Currency of Science (Aarhus University, evaluated with top grade; currently being polished for submission).
- Master’s thesis (University of Oslo): Ergonomics of Discourse: Measuring Interpreter Load Indices in Human- and Large Language Model-Generated Text (in revision/polish for submission and publication; submitted as a talk to SALC 2026).
Selected applied/commercial research (Gartner)
- “Uncover the Hidden Impact of AI Automation on Your Organization”, ID G00794951 — Gartner (by Jo Karajanov, Mary Mesaglio)
- “How Executive Leaders Can Reframe Bias With Generative AI”, ID G00793053 — Gartner (by Mordecai, Jo Karajanov)
- “9 Trust and Ethical Implications of Generative AI”, ID G00795391 — Gartner (by Bart Willemsen, Jo Karajanov)
Current writing interests
- Cognitive ergonomics, interpretive load, and discourse-level meaning
- Human–AI interaction as a cognitive ecology
- Epistemic agency and responsibility under fluent automation
- Governance language, organizational narratives, and measurable rhetoric
Teaching
Teaching philosophy
My teaching is practical, concept-driven, and respectful of the learner: I treat methods as tools for thinking, and gamify wherever possible. By “gamify” I do not mean the insistence on points and levels, and instead mean that I respect the role of (obstacle course) design and meaningful effort. Gamification, in this sense, is about making the abstract knowledge students need to attain a tool to be used in the pursuit of a tangible, recognizable goal. I aim to reduce intimidation, increase agency, and build durable intuition—especially around statistics, coding, and research design.
University of Oslo
- Teaching Assistant — Introduction to Statistical Analysis for Language Students (R) (Autumn 2024; Autumn 2025)
- Taught statistics and coding in R to language students
- Created synthetic datasets and authored learning games (including a “Forensic Linguistics Lab”) to support applied learning and retention
Earlier teaching (selected)
Lecturer — Academic English & Research Methodology for Game Developers (Technische Universität Köln / Cologne Game Lab). Topics included game theory, cognitive biases, immersion, creative co-construction, and teamwork in non-linear systems.
Experience
Current / recent roles
- Executive in Residence — NLP & Data Analysis, BI Norwegian Business School (Oslo). Research into the language of annual and sustainability reports; corporate governance disclosures and structures; published output in MAGMA.
- Co-founder / daglig leder / researcher — Kanna AS (Oslo). An independent research vehicle enabling focused project work and research collaborations (mostly paused while studying).
- Senior Director, Analyst — Gartner (Berlin/Munich/Oslo). Research, analytics, and advisory on machine learning and generative AI; publication-quality applied research outputs.
Higher education leadership
- Chief Information Officer — Frankfurt School of Finance & Management (Frankfurt am Main). Led transformation strategies toward data-driven management; cloud migration; designed and directed an AI Initiative and opened an AI Lab.
- Academic Programme Director — MSc Applied Data Science. Designed, staffed, and recruited for the program and achieved accreditation in the first year.
Earlier professional work
I also worked as an in-company business English trainer across multiple large organizations in Germany, focusing on professional communication and practical skill-building.
Earlier creative/industry work (a bit of trivia)
Voice-Over Direction (Game Development). Before moving fully into higher education leadership and research, I worked as a Voice-Over (VO) Director in game development. The work combined narrative editing, casting calls, and directing voice actors remotely.
Volunteering & Academic Citizenship
- University of Oslo — Studentrepresentasjon i UHRs nasjonale fagorgan / Student representative in UHR's national academic councils: HFM2-LING Lingvistikk (October 2025 – Present)
- University of Oslo — Member, “Kunstig intelligens i analyse av språk og språkdidaktikk”, ILN (January 2025 – Present)
- Oslo Fireside Knits — Organizer (Jan 2024 – Present). Community knitting group focused on slow fashion and sustainable garment making; designed and delivered a curriculum covering traditional knitting skills.
- Venturing Women Lab — Technical Mentor for a short course (March 2024 – May 2024)
Contact & Connect
- Email: jovanak@uio.no
- ORCID: To my ORCID page
- LinkedIn: To my LinkedIn page
© Jovana (Jo) Knutsen. Built as a lightweight static site.