This book is concerned with a family of current and recent debates which may be loosely gathered under the title of realisms versus anti-realisms. Wright's first aim is to establish a set of theses which should be by all parties to the various debates-though they have not been so in fact. Setting aside these agreed theses, Wright proceeds to sketch some of the cruces at which, he thinks, realists and anti-realists should properly part company. Some of Wright's cruces are familiar, but some are radical reworkings of familiar cruces. In short, Wright proposes to generate a new agenda for debates about realism. Finally, Wright conjures up a spectre from Wittgenstein's rule following considerations which threatens to produce an outright victory for anti-realism, and Wright tries, tentatively, to contain this threat. The anti-realisms most discussed include theories of moral discourse (locus classicus John Mackie 1977) and of mathematical discourse (locus classicus Hartry Field 1980). Error theorists have held that our ordinary moral and mathematical assertions are apt for being or false, in a sense of true which is not merely disquotational but supports the idea that statements correspond to the moral or modal facts. But they claim that such statements are in fact false, because the world does not contain what would be required to make them true, and hence we are in massive and systematic error in our moral and mathematical beliefs. Other anti-realist theories in view are expressive, non-cognitive, projective accounts of moral or modal discourses, which claim that moral or modal utterances are not properly regarded as being apt for truth or falsity, if truth is not merely disquotational but supports the idea of correspondence to moral or modal facts. Also in the picture is the debate between a Dummettian realist, who claims that our grasp of truth conditions is such that truth is not constrained by possible evidence, and his opponent, the Dummettian anti-realist. The list of anti-realisms is left open-ended. Wright argues that all these various anti-realists should agree, pace expressivists, non-cognitivists, etc., that the target discourse, moral discourse or modal discourse for example, is apt for truth or falsity, where the concept of truth concerned is not merely disquotational but expresses a genuine property which