This paper argues a number of points about measurement in the sociology of mental health: (1) measurement is critical, (2) measures should represent and assess elements of human experience, taking measure of life as people feel it, sense it, and understand it, and (3) social scientists should create a human science, producing information for the people it studies so that they can better understand and control their own lives. We argue that a human science is best achieved with the use of indexes, not diagnoses, to measure mental health. We present a brief history of diagnostic instruments and detail how a diagnosis is made. We show how use of diagnoses to measure mental health discounts much human suffering. They dichotomize the true range of feelings and emotions into crude either/or distinctions that do not reflect the reality of people's lives, and they often exclude suffering such as that due to loss or illness that does not meet medical model preconceptions about mental disorder. Using diagnoses to measure mental health presents a reified image of hidden disease knowable and manageable only by trained professionals--beyond the capacity of the suffering individuals to understand and control.