This thesis is an enquiry into expert-intensive work to support the aggregation and mass-dissemination of scientific articles. The enquiry draws on computer-supported actions and interactions between the people who worked as daily operators, and submitters who worked the system from remote as end-users. These recorded actions and interactions render visible matters of practical responsibility and competence with reference to reportable-accountable use of objects and devices. As a scholarly contribution, this thesis draws on the work started by Garfinkel and his followers, ie., the study programme of ethnomethodology. It poses a simple question—how do the people involved in these computer-supported actions and interactions do just what they do? In the attempt to answer that question, the enquiry shows that a competent and accountable use of system supports is manifested in reference to particular phenomena that need constant management—eg. warnings,anomalies, unknown entities, failing processes, disorderly work objects, and spurious actions—in short, phenomena of disorder. Records of how these phenomena are detected and attended to, and how particular problems are solved, reveals a complex relationship between computational functions and subtle human judgements. Without making generalised theoretical claims about this relationship, I conclude that remarkably little is known about the actual lived work that takes place in the operation and use of computer systems that are worked specifically to detect and cope with `anomalies' and thus require detective and reactionary labour. I offer this enquiry into the goings-on at this particular site as a singular opportunity to respecify problems of disorder.