In this paper we suggest that discussions about the identity of recipes should be based on a distinction between four categories of recipes. The central feature that we use to single out a category is the type of relationship that a recipe bears to its author. The first category comprises “open recipes” like wine, pizza, or salad, which come in taxonomic layers and are structurally open for new authors to reshape them. The second category comprises “institutional recipes,” namely those whose (...) authors typically form consortium-like institutions, such as Champagne wines or Quebec maple syrup. The third category comprises “brand recipes” like Coca-Cola, Nutella, or Big Mac, whose names are elusive semantic devices and connote rather than denote recipes. Finally, the fourth category comprises “flagship recipes,” which include all the personal renditions of a recipe whose identity is strongly bound to individual authors; we suggest that their semantics follows a causal-reference model of proper names. Besides its theoretical value, the classification we put forward is offered as a ground for settling legal disputes about recipes, evaluating charges of cultural appropriation that concern recipes, and guiding consumers, producers, and policy makers when they think about foods and diets. (shrink)