This chapter draws on a study of amateurs' -music-and food-lovers' -practices, to show that taste is an activity and not a passive or determined state. We use the words 'amateur', 'taste' and 'lover' in their broad senses referring to any form of love or practice, and not only the restrictive cultured sense of a connoisseurship centred on a knowledge of the object itself. Amateurism is contrasted, on the one hand, to the lack of concern of lay-people who pay little attention to what they eat or listen to, and, on the other hand, to the certified expertise of the professional. These three levels differ more as types of engagement than as degrees of intensity. We were especially interested in great amateurs, not because their knowledge of the product is greater, but because their reflexive activity on the object of their passion more clearly reveals the diverse forms and devices, gestures and timings, training and guides, needed for such an involved taste to develop. Throughout the discussion we also examine systematic comparisons between types of attachment, contrasting food with, primarily, music: a historical repertoire and a complex performance, centred on a mysterious state of emotion induced, in the case of music; and, by contrast, a concentrated object, giving immediate physical pleasure or satisfaction and instantly destroyed by its own tasting, in the case of food. The idea was the same: comparison is a good means by which to further our understanding of these diverse forms and media of taste, depending on the products tasted.Our purpose here is to present and justify a new research programme on food taste. A systematic and critical review of the most prominent research on the subject has revealed that when it comes to the status of products concerned with taste, the various disciplines are divided, unsatisfactorily, along the lines of a nature-culture approach: either food products are just things and their properties are analysed through laboratory tests and measurements; or they are simply signs, the media for various rites and mechanisms of social identity, in which case their physical reality disappears in the analysis. In our opinion, this duality is detrimental in so far as it eliminates the very object of taste, i.e. it obscures the inevitable uncertainty about the effect that arises, about