Distributed systems are getting more and more numerous, complex and used in a wide variety of applications. New solutions and new architectures arise (e.g., clouds) that support new functionalities (e.g., social networks). They pile up several software layers and, given that any software is directly dependent of the underlying layers, it can be unable to provide promised services whether any of these layers misbehaves. This evolution implies new non negligible dependences increasing with the number of actors involved in the system (e.g., providers and users). Some dependences could be hidden by this layer stacking, implying a reduced transparency for users and a misunderstanding of her actual autonomy. We argue that users should be aware of the potential risks resulting from these dependences. To be able to deduce them, one should know the way the system works (architecture, involved resources, providers, participants, etc.). This would help to deduce the potential trust a user could or should have toward the system. We consider this of utmost importance, as professional efficiency and personal privacy could be compromised if untrusted actors control the access to others' resources. This work proposes SOCIOPATH, a generic meta-model that allows to expose hidden or implied relationships among participants in the digital world, which also introduce dependences at the social level. The notions presented in this approach are basics of many fields, like security, privacy, trust, sociology, economy and so forth. SOCIOPATH can be used in the evaluation process of a system as well as in its upstream design.