Abstract
Initially found in population studies designed to discover a link between child abuse and population categories, risk has been institutionalized in British Columbia through the use of a risk assessment tool presumed to measure danger to particular children. Recruitment of the risk speech genre reflects a need for government child protection workers to clearly articulate which children are in need of protection from “risks as they really are” while avoiding the accusation of “intervening too much.” Moreover, risk assessment tools are both audit systems and auditable systems connected through an audit chain. Auditors who represent potential blaming procedures track errors in risk valuation to specific decisions. This entrenches a separation between what we want to know about risk from what we want to do about risk. The apparently objective use of numeric values within risk assessments obscures the continuing role of subjective moral values in the practice of child protection.