nonmonotonic reasoning

Nonmonotonic reasoning is a type of logic where conclusions can be revised or canceled in light of new information. Unlike traditional monotonic logic, where adding premises never invalidates existing conclusions, nonmonotonic reasoning acknowledges that new evidence might require a change in belief.1

This is crucial in artificial intelligence and other fields because real-world information is often incomplete. For example, knowing that "Tweety is a bird" and "birds fly" might lead to the conclusion that "Tweety flies." However, if new information emerges, such as "Tweety is a penguin," the initial conclusion must be revised.1

This self-correcting principle is fundamental to how intelligent systems can adapt and learn from evolving data. It differs from defeasible reasoning, where inferences themselves can be blocked or defeated by new beliefs.1