Abstract
I have recently argued in various places that each kind (as the word is defined in the paper) has a history, and only kinds have histories. Because 'art' is shown not to be the name of a kind it does not name a kind with a history. This conclusion relies upon a meme-driven account of cultural evolution, providing historical studies in general with a rationale running intellectually parallel to that which sustains objective accounts of biological evolution. I propose that the word 'art' should not be used in the conventional way but as a name for the general category of memetic innovation, and the phrase 'work of art' as a name for the items of the class of entities inspected by an unprompted eye in search of memetic innovation. These arguments are recapitulated in a form encouraging speculation about their consequences for the art museum and gallery. One effect of a conceptual revision would be to widen the scope of the art museum: on the face of it, without restriction. In order to impose a practical constraint it is suggested that an institutional criterion might be applied to acquisitions and display policies, simultaneously and consistently with a rejection of the Institutional Theory of Art that is currently influential in the art world. It is further argued that the art gallery should be conceived primarily as a domain of entertainment with the underlying function of encouraging creative discovery by the unprompted eye, and not as an art-historically instructive domain. The art museum's collection will inevitably lend itself to study and presentation in various intellectually coherent patterns by cultural historians, but-if the thesis is correct-not by those 'art historians' who mistakenly suppose that a bogus 'history of art' provides an appropriate structure for the understanding of works of art.