Abstract
The more recent philosophical literature concerned with foundational questions about normativity often appeals to the notion of normative reasons, or considerations that count in favor or against actions, and their interaction. The interaction between reasons is standardly conceived of in terms of weighing reasons on (normative) weight scales. Knoks and van der Torre [8] have recently proposed a formal framework that allows one to think about the interaction between reasons as a kind of inference pattern. This paper extends that framework by introducing and exploring what we call numerical balancing operators. These operators represent the weights or magnitudes of reasons by means of numbers, and they are particularly well-suited for capturing the intuition of aggregating and weighing reasons. We define a number of concrete classes of balancing operators and explore them using a principle-based analysis.
Aleks Knoks
PostDoc Researcher
Currently, I’m mostly focused on two projects. The first uses formal tools to make progress on some foundational questions in (structural) metaethics, as well as on the challenge of designing artificial agents that can act in ethically acceptable ways. The second project is concerned with epistemological questions surrounding black-box AI systems and the field of explainable artificial intelligence (XAI). My present position at the University of Luxembourg is funded through the project the Epistemology of AI Systems (EAI), which I co-wrote.
Leon van der Torre
Professor of Artificial Intelligence and Computer Science
Leon van der Torre is a full professor of artificial intelligence and computer science at the University of Luxembourg. He has been working on deontic logic and its applications for nearly three decades. He is one of the founding fathers of input-output logic, the BOID architecture, the game-theoretic approach to normative multiagent systems, and the LogiKEy methodology.