How to solve the conjunction fallacy? A discussion of alternative approaches Dewi S. Harten Abstract: How to solve the conjunction fallacy? A discussion of alternative approachees Dewi S.Harten Abstract: Since many years, people have tried to understand and to model their own behavior and language. This has proven to be a difficult, yet rewarding task. Since the development of artificial intelligence and robots, there is an actual use for this modeling, besides the sheer happiness of understanding a part of our own existence. One of the problems that we encounter when dealing with human language, is our way of handling concepts and creating new, complex concepts out of simpler ones. A concept can be seen as a mental representation [the Stanford Encyclopedia of Philosophy, 2007]. For example, let us consider a real life object ‘chair’. This is something different from what we have in mind when we say the word ‘chair’. Because when we are talking about it, we are using a mental representation and not the object itself. In other words, our mental representation will probably be a prototypical chair, where the real life chair does not need to be. When we are creating complex concepts out of simpler ones, we are actually trying to describe the complex object we see, by combining two or more simple concepts that are already familiar to us. This thesis will investigate the different possibilities to solve one of the problems that occur when we are trying to create complex concepts: the conjunction fallacy. The conjunction fallacy occurs when the combination of two concepts has a higher probability than the original concepts. This thesis will explore what research has been done through the years in this field. It will define different ways in which the fallacy can be interpreted and it will try to find a solution for the conjunction fallacy. I have divided my thesis into three parts. The first part handles the different approaches to a solution for the conjunction fallacy using a ‘classical’ Boolean algebra. The second part handles the more recent approaches that use a non-Boolean algebra and geometrical models. Finally, the third part contains the conclusion and future work.