Deliberative democracy is based on the belief that, under the right conditions, talk-based conflict resolution can be a powerful way of solving social and political problems. This belief is in turn supported by large bodies of theoretical and philosophical work as well as a great deal of empirical research. In both theory and practice, however, deliberative democracy has run up against "tough cases": situations in which identities and interests are so polarized, or the sources of conflict are so severe, that the necessary conditions for deliberation are missing. The purpose of my research is to develop the tools deliberative democrats need to deal with these cases, starting from the premise that in situations of conflict, instead of an interest in solving the problem through discussion, people prioritize their personal interests or those of their group. There is a tendency in deliberative democracy to assume that the strategic pursuit of these interests undermines deliberation, and that the best way around this problem is to eliminate, as far as possible, opportunities and incentives for people to act strategically. However, the worse the conflict, the more difficult this approach becomes, making it unreliable for tough cases. Luckily, it is not the only way of looking at the problem, nor is it necessarily the best. This research will offer an alternative: a theoretical and practical account of how strategic incentives might be channeled into deliberation so that people's own interests might support, rather than undermine, talk-based conflict resolution in a wide array of cases, from ethnic, cultural or religious conflict to highly polarized debates that pit economic welfare against environmental integrity.