One leading way to study consciousness empirically is through its links with metacognition. Indeed, some theories of consciousness equate them: conscious states just are metacognitively monitored states. The other leading scientific theory of consciousness is the global workspace theory. These two approaches are usually taken to be rivals.
That is a mistake. In a paper in Trends in Cognitive Sciences, Chris Frith and I argue that they are in fact complementary. There are strong reasons to expect them to go together. Given the function of a global workspace, we should expect representations that are selected and processed so as to be made available to the global workspace to contain a metacognitive component; or to be accompanied by a metacognitive content, such as a rating of high or low confidence in the content being distributed.
As well as tracing through this functional argument, we review the existing empirical evidence for and against our hypothesis, and suggest ways that it can be further tested empirically.