Toggle Main Menu Toggle Search


On the Trustworthiness of Computing Systems

Lookup NU author(s): Emeritus Professor John Dobson, Professor Brian Randell



A definition of what we here call ""trustworthiness"" is proposed as an extension of the concept of dependability, which makes trustworthiness a judgement about the acceptability of a system rather than a property of a system. It is argued that this approach is appropriate to those large and complex systems where the rich diversity of human roles associated with the system means that the various expectations of the system cannot be reconciled into a single authoritative specification, and therefore a system can, in human terms, be judged a failure even when it behaves in conformance with some (presumably previously agreed) dependability specification. As a consequence of treating failure as subjective and trustworthiness as a judgement, we investigate new logical bases for discussing such things as the interpretation of evidence for trustworthiness, the predictability that a certain judgement will be made, the role(s) of the judge(s), the definition of system boundaries, and the criteria for acceptability.

Publication metadata

Author(s): Dobson JE, McDermid J, Randell B

Publication type: Report

Publication status: Published

Series Title: Computing Laboratory Technical Report Series

Year: 1990

Pages: 28

Report Number: 306

Institution: Computing Laboratory, University of Newcastle upon Tyne

Place Published: Newcastle upon Tyne