Abstract
The potential benefits, costs, and risks of severe violence dialog (SVD) are analyzed using constructivist theories. Implications for mediating SVD, selecting and preparing participants, securing informed consent, and controlling the process are discussed. This paper examines the need for more sophisticated guidelines for SVD, and for scientific research aimed at improving participant selection and preparation, a fuller understanding of SVD dynamics, and better control over the process.