We are confident of many of the judgements we make as to what sorts of alterations the members of nature’s kinds can survive, and what sorts of events mark the ends of their existences. But is our confidence based on empirical observation of nature’s kinds and their members? Conventionalists deny that we can learn empirically which properties are essential to the members of nature’s kinds. Judgements of sameness in kind between members, and of numerical sameness of a member across time, merely project our conventions of individuation. Our confidence is warranted because apart from those conventions there are no phenomena of kind-sameness or of numerical sameness across time. There is just “stuff” displaying properties. This paper argues that conventionalists can assign no properties to the “stuff” beyond immediate phenomenal properties. Consequently they cannot explain how each of us comes to be able to wield “our conventions”.
Elder, Crawford, "Conventionalism and the World as Bare Sense-Data" (2007). Articles. 8.