<p>Many social policy problems are recognised as complex and intractable, and hence necessitate analysts' having the capability to address them. Epistemological influences embedded in approaches to policy can impose constraints on the natural capacity and capability that people have to make sense out of particular experiences of complexity in the course of policy analysis work. Within the dominant policy approach adopted by policy analysts under the rubric of evidence-based policy, such complexity capability eschews any explicit role for opinion. However, the application of Q methodology by Michel van Eeten among others in a specific case of policy deliberation in the Netherlands, which had proven resistant to the standard, evidence-based policy analysis, shows that there could be a role for what is otherwise overlooked. Accordingly, this thesis examines the proposition that opinion indeed may play an important role in policymaking in complex and intractable situations. Q methodology is an established research methodology for acquiring and developing knowledge from a subjective standpoint. It has a growing record of successful application to public policy controversies, where solutions were made possible because opinion - and its everyday experiential rationality - were made available. Q methodology is also seen, however, as a marginal methodology. There has been insufficient explanation of why the application of Q methodology could make a positive difference to policy problems of a complex and intractable kind. The two research questions focus on the efficacy of Q methodology. Q methodology could make a difference in an adjunctive sense. It meets a policy need, namely to make opinion available as a complement to other evidence knowledge and thus adds to understanding of problems and solutions while remaining firmly within the prevailing evidence-based policy epistemology. Alternatively, Q methodology could make a difference of a transformative kind. It opens up a new epistemological space for doing policy analysis work with the power to create substantial policy-analytic change. To address these questions, the thesis develops an argument that establishes the linkages between pragmatism, complexity thinking and Q methodology and, in so doing, provides a path for understanding the role and place of opinion in policy making contexts. It proceeds through several stages which together make an epistemological argument for the efficacy of Q methodology. First, the nature of the policy problem is explicated as one of the separation of opinion from knowledge. Secondly, the thesis turns to a counter argument drawing on Peirce's pragmatism and his attention to abduction. In the next stage, dominant practice ideas about the capability needed to address complexity are critically examined, which shows that opinion is not valued in that practice. The success of van Eeten's work leads to a detailed examination of complexity in the policy context, and the claim that opinion is less problematical than are the overall epistemological choices made in policy analysis. Focusing on those epistemological choices, the argument draws together, in a fresh look, the thinking entailed in Q methodology in respect of its abductive logic and its theory of knowledge. Q methodology is shown to be a kind of science that allows objective fact to be approached from a subjective standpoint under experimental conditions. Finally, therefore, Q methodology is shown to open up an epistemological space quite unlike others. This makes the practice described as "reading complexity" in a real-world policy application possible.</p>