What I like to call the problem-solution paradox states that we cannot think about solutions until we understand the problem, and we cannot understand a problem until we think about solutions.
    The first part of the paradox is familiar. Good design processes attempt to clearly define a problem space before designing solutions, thus cutting down on wasted time and resources. Designers move linearly from understanding to framing to ideation to validation, ensuring rigor at each step to inform the next. However, there are issues with this linear course. First, it assumes that there is a “final answer” at the end of the “understanding” phase, and, once we find it, we can design for it. This would be fine if design were an analytical task in which all problems have solutions. But as we have seen, design is abductive in the sense that designers are not looking for Truth but rather the many truths that might be appropriate to specific contexts. We should, therefore, resist the urge to seek final answers. Second, the linear approach assumes that designers are able to understand a problem space as divorced from potential solutions. Although this might be true in part, it does not account for the complexity of behavior. The linear movement from problem to solution does not leave room for exploring the effects of solutions on the system for which they are designed. It is a theoretical stance—even with the most rigorous contextual research, most findings are theoretical in nature. 
    The second part of the paradox attempts to solve these issues, but does so by simply reversing them. It allows the exploration of possibilities via prototypes, but in the process it eliminates valuable insights that come from theoretical work. Using solutions to understand a problem space is a purely practical approach—the creation of entities in a system elucidates the nature of that system. While this approach sounds good on the surface, it can easily be interpreted as a means to eliminate the complex parts of design: observing complex behavioral systems, and understanding cognitive patterns beyond what someone consciously says.