Judges on a multimember court might vote in two different ways. In the first, judges behave solipsistically, imagining themselves to be the sole judge on the court, in the style of Ronald Dworkin’s mythical Judge Hercules. On this model, judges base their votes solely on the information contained in the legal sources before them—statutes, regulations, precedents, and the like––and the arguments of advocates. In the second model, judges vote interdependently; they take into account not only the legal sources and arguments, but also the information contained in the votes of other judges, based on the same sources and arguments. What does the law say about these two models? May judges take into account the votes of colleagues in deciding how to vote themselves? Should they do so? Are there even conditions under which judges must do so?