The emergence of the IoT as an everyday fact raises the question how the IoT can be trusted. Considering that the IoT is pervasive to the level of being secretive, practically monopolistic and that human participation is often involuntary, it hardly satisfies the assumptions that associate the often researched types of trust relationships. In order to study trust in the IoT, alternative views on trust may be needed. This paper analyses the IoT as a representation of imperfect systems, i.e. systems that by architectural choices are unable to guarantee predictably repeatable operations. This property may invalidate the metaphor of trusting technology that is constructed out of replicating human-to-human trust. This paper examines alternative views on trust that may better fit the specificity of the IoT and generally imperfect systems, adopted from psychology, sociology or ergonomics. While no definitive approach is indicated, this paper serves as an overview of possible directions in trust research.