The Diagram Puzzle
The goal of the paper is to argue against the claim that thoughts can be modelled as having a diagram-like structure. The argument has a form of the so-called Diagram Puzzle, according to which the same features make diagrams cognitively reliable (and desirable) and unreliable (and non-desirable). I argue that to solve the Puzzle we have to accept the instrumental interpretation of diagrams, according to which diagrams are instruments of reasoning comparable to calculators. Instrumental view on the nature of diagrams leads to the problem of content determination: the claim that instruments can determine thoughtsâ€™ content, entails that, for example, a calculation carried out with fingers has a different content that the same calculation carried out with abacus. If instruments do not determine content, they can be seen as instruments that reveal the content of thoughts, but they do not change the thoughts content. I argue that diagrams are epiphenomenal which means that they cannot influence the thoughtâ€™s content. Therefore, we can think with the help of diagrams, but it does not follow that thoughts have a diagram-like nature.