Despite its pervasiveness, the concept of ‘levels of organization’ has received relatively little attention in its own right. I propose here an emerging approach that posits ‘levels’ as a fragmentary concept situated within an interest-relative matrix of operational usage within scientific practice. To this end I propose one important component of meaning, namely the epistemic goal motivating the term’s usage, which recovers a remarkably conserved and sufficiently unifying significance attributable to ‘levels’ across different instances of usage. This epistemic goal, to (...) provide structure to scientific problems, delegates tasks whose execution generates the term’s expressed content in a given instance. This treatment of levels does not diminish the concept’s general importance to science, but rather allows for its use in, and usefulness for, scientific practice to be better contextualized to particular tasks encompassing varying breadths of activity. (shrink)