Richard Rorty Consider maps. Does a landscape tell us how it is to be mapped? In one sense, clearly not. Your purposes may dictate that you map on any of many scales, depicting topography, population, rainfall, geology, or many other things. Here you can choose whatever turns out useful. You can stress what you like and be as vague or precise as you like. Sometimes a brief sketch will do, and sometimes only an admiralty chart. Pragmatism and Darwin and multiple perspectives are all in order. There is no competition between a geological map and a rainfall map. But in another sense the landscape indeed dictates something. It dictates how it is to be mapped, given a set of conventions determining the meanings of the signs and shapes on the map, and the meanings of their presence or absence. That is why, once a set of conventions has been put in place, a map can be correct or incorrect, or in other words, how it can represent the landscape as it is, or represent the landscape as it is not. It can show cliffs where there are none, and fail to show cliffs where they lurk. These platitudes should be distinguished from the ludicrous idea that the only true map would map the landscape an sich, somehow embodying a "final vocabulary" or "nature's own vocabulary" dictating how it is to be mapped, as if human selections and purposes had nothing to do with it.