Philosophers of law have spilled oceans of ink analysing how rules work, how they are read from the text, and how they make up larger systems called institutions. The knowledge gained in this way can be useful to those who analyse rules in political science. In this paper, I discuss several insights from legal philosophy that can be used to rethink the way rules are usually analysed in the social sciences. I will show the differences between deriving a rule from linguistic signs (such as a text) and natural signs (such as the behaviour of other people), analyse the complex interplay between a rule and the practice of applying it, and explain why the inference often made in the social sciences about the similarity of two rules or institutions on the basis of similarities in the text is fraught with a high risk of error. The insights of legal philosophy are also helpful in analysing the evolution of rules, their path dependency, and how hard cases affect the history of their application. The paper concludes with some recommendations on what to do and not to do when measuring rules.