Cognitive Science Colloquium | Universität Tübingen

Apart from the Cognitive Science Colloquium, there are also many other Cognitive Science related talks / talk series in Tübingen that are listed here.

Abstracts

27.10.2026 Between Habermas and Hate Speech: Studying Human Behavior in Comment Sections

Speaker: Prof. Jürgen Buder, Leibniz-Institut für Wissensmedien, Tübingen

In the 1990s, many scholars believed that exchanging opinions in online forums and comment sections would be the ideal Habermasian public sphere where citizens may engage in rational discourse on societal issues. 30 years later, little is left of this idea: Social media are widely seen as venues for hate speech and hotbeds for societal polarization. This talk introduces psychological research on human behavior in comment sections. Experimental studies and two large-scale field investigations show that humans preferentially engage with dissenting others in comment sections. From the standpoint of deliberate democracy, that is good news. The bad news is that this clash of viewpoints may increase rather than decrease polarization. Current and future directions of this will be discussed.

08.12.2026 Talks by Cognitive Science PhD students (University of Tübingen): Title 1: How sound context shapes perceptual timing precision; Title 2: Motor Metacognition in Music Performance

Speaker 1: Charlotte Mock (Cognitive Neuroscience of Human Verbal Communication, Universität Tübingen)

Abstract 1: Abstract: In everyday listening, temporal structure allows listeners to predict the timing of upcoming sounds. In speech, such predictions may support the analysis of temporal variation relevant for segmentation, prosody, and emphasis. To use temporal information effectively, listeners must perceive the timing of sounds with high precision. But what determines this perceptual timing precision (PTP)?
Previous research has largely attributed differences in PTP to the acoustics of individual sounds. Here, I will present behavioral evidence that PTP critically depends on spectrotemporal continuity between a target sound and its preceding auditory context. Finally, I will propose a biologically plausible computational model that captures our behavioral findings, providing a potential mechanistic explanation for how acoustic context shapes timing precision.
Together, these findings suggest that auditory timing precision is determined not simply by the acoustic properties of individual sounds, but by their relationship to preceding auditory context, offering a new perspective on temporal prediction in speech and other complex auditory environments.

————-

Speaker 2: Maria Paula Villabona (Metamotorlab, Universität Tübingen)

Abstract 2: Humans can not only decide, perceive and remember, but also reflect on how well they do so. This capacity, known as metacognition, is thought to offer a wide range of functional benefits, especially for learning and achievement in educational settings. Its role in motor performance, however, remains less clear. Skilled movement is often viewed as operating largely outside conscious awareness, yet recent work suggests that performers have measurable access to details of their own movements. Whether, and under which conditions, this monitoring ability contributes to motor performance remains an open question.

In my dissertation, I investigated the role of metacognitive monitoring in motor performance, using guitar playing as an ecologically valid task. I will present results of a study where we tested whether monitoring ability moderates the Focus of Attention (FOA) effect, which states that an external focus (i.e., on the effects of one's movements) leads to better motor performance than an internal focus (i.e., on the movement itself). I will present analyses based both on participants’ responses to a cognitive task, and on their movement trajectories.

Together, these results present evidence to move beyond the idea that metacognition uniformly supports motor performance, toward a more precise mechanistic account of the circumstances in which metacognitive monitoring influences skilled behavior, with possible implications for how motor skills are learned, evaluated and regulated in real -life contexts.

15.12.2026 Minds Measuring Minds: Can We Trust Our Measurements? Reliability and inter-observer agreement in Cognitive Science

Speaker: Sophie Vanbelle (University of Masstrich)

How do we transform something as complex as the human mind into a number? Cognitive scientists routinely measure constructs that cannot be observed directly, using questionnaires, behavioural tasks, physiological signals, brain images, and increasingly algorithms and AI. Yet these measurements are not the constructs themselves: they only are imperfect representations. If two researchers, instruments, or algorithms assess the same phenomenon, should we expect them to agree, and what does it mean when they do not?

This talk introduces the statistical ideas behind reliability and agreement without assuming an advanced statistical background. Using examples from psychological and health research, I will show how different measures of inter-observer variation answer different scientific questions, with consequences for study design, uncertainty, and the interpretation of results.

I will then turn to my current research. As cognitive science moves towards repeated and high-dimensional measurements, brain imaging, and AI-generated predictions, classical statistical methods face new challenges. I will discuss how we can extend methods for evaluating agreement and reliability to these settings, and why, even as producing numbers becomes easier than ever, inter-observer variation and reliability remains just as important.

12.01.2027 Cognitive Offloading in Human-LLM Interaction: From Addiction Bias via Memory Costs to Scaffolding

Speaker: Dr. Frank Papenmeier

People increasingly rely on Large Language Models (LLMs) to perform cognitive tasks. From a psychological perspective, these models can be investigated in two distinct ways: as artificial agents whose behavior is evaluated experimentally, and as cognitive tools that alter human information processing. First, treating AI as a subject of psychological research, I show that LLMs systematically replicate and amplify a human heuristic, specifically the addition bias. Second, shifting to the human side, I use the framework of cognitive offloading to examine the determinants of offloading mental effort, including bot response speed and user ability, and demonstrate that relying on AI solutions incurs substantial memory costs during subsequent recall. Finally, to counter these costs, I present an LLM tutor designed as a cognitive scaffold. By constraining the system to provide adaptive guidance rather than direct solutions, this setup supports skill acquisition. Together, these findings show that whether AI impairs memory or fosters learning depends on how the interaction is structured.

19.01.2027 Automated Scientific Discovery of Mind and Behavior

Speaker: Prof. Dr. Sebastian Musslick

Artificial intelligence is transforming scientific discovery across disciplines, from predicting protein structures in biology to identifying novel quantum states in physics. In the natural sciences, automated discovery systems now enable researchers to navigate vast experimental and hypothesis spaces, moving beyond the limits of human intuition. In this talk, I explore how these methods can be adapted to the study of mind and behavior, introducing automated scientific discovery as a paradigm for cognitive science. I present AutoRA, an open-source framework for automating key stages of empirical research -- including experimental design, data collection, and model inference -- and demonstrate how it can be used to uncover new computational models of cognition and novel behavioral phenomena. Through case studies in psychophysics, learning, decision-making, and cognitive control, I illustrate how closed-loop discovery systems can support more integrative, scalable approaches to understanding human cognition. Finally, I discuss the unique challenges that automated scientific discovery faces in cognitive science, such as experimental fragmentation and the need for multi-level explanations, and outline future directions for building systems that not only accelerate scientific progress, but reshape how we generate and evaluate theories of the mind.