Philosophers have recently been interested in the appropriate attitude to have toward a controversial philosophical theory. There are a variety of reasons to doubt that belief is appropriate, as the high epistemic standards required for justified belief would seem to exclude too many philosophical theories as viable options. Relatedly, there are worries about assertions of philosophical claims, since assertion is also thought to be governed by substantive epistemic norms. In light of this, a variety of alternative attitudes have been proposed in order to characterize philosophical commitment and assertion. 1 Alexandra Plakias (2019) has helpfully expanded this discussion to the topic of philosophical publication. Plakias argues that it is permissible for philosophers to write and publish works (e.g., journal articles or books) that include claims they do not believe, a practice she calls publishing without belief (PWB). 2 Her thesis is carefully limited to philosophy, as opposed to publishing in other fields. Plakias' appeals to three helpful cases to provide a compelling argument that PWB is sometimes permissible. However, this raises further questions: just when is PWB permissible? Is it universally or even generally permissible? I argue that only claims of a certain kind are permissible to publish without belief. The picture I propose recognizes two roles for claims made in published works: what I will call the advocacy and evidential roles. Advocacy claims should often be published despite authors' lack of belief in them. Evidential role claims, however, require a higher epistemic standing, one that does include a belief requirement. On this theory, Plakias' cases are permissible examples of PWB because they involve advocacy role claims. I will offer another three cases illustrating that it is impermissible to publish evidential role claims without belief. The resulting account of the norms of publishing is applicable to research areas beyond just philosophy.