This paper presents a policy-conflict handling model, which is independent of both technical details of policy enforcement architecture and syntax of policy-specification language. The model primarily focuses on handling the application-specific (semantic) conflicts [1]. This model has been applied to the POLICE Policy-Based Network Management (PBNM) framework in [2-4]. The POLICE framework does not suffer from the modality conflicts [1] since it does not support negative policies whereas application-specific (semantic) conflicts are handled according to the model described in this paper. We are able to rely on a common schema in our model since almost all policy specification languages include subject, target and action concepts and the fact that all languages can be modeled in an approximately similar way allows our model to be designed independently of other details of the language.