Suicide often occurs unexpectedly, both for clinicians and close others. Despite decades of empirical research, meta-studies show that the predictive accuracy of known psychological risk factors remains only marginally better than chance when applied to individuals. This paper addresses the pressing question of what it means to expect suicidal behavior when an individual's reasons for living (RFL) and reasons for dying (RFD) are known. Drawing on tools from philosophical action theory, I propose to understanding suicidal behavior not in terms of statistical likelihood, but in terms of explanatory coherence. On this view, suicidal actions are unexpected when they are difficult to explain in light of an agent’s RFL and RFD. The paper critically examines index-based psychological models of suicidality and argues that they miss the normative and evaluative structure of reasons. I argue that Alan H. Goldman’s theory of well-being offers a promising basis for understanding suicidality. Suicidality becomes intelligible when life is missing the very components that typically bind us to life, which are the key elements of well-being themselves. Rather than assigning weights to reasons for living and dying, I propose to examine the intelligibility of patterns across well-being dimensions, which reveal whether suicidal actions are evaluatively coherent: the less coherence, the harder the suicidal action is to explain, and thus the more unexpected it is.
René Baston (Tue,) studied this question.