It remains unclear to me why you are skeptical of rational agency despite having no problem with rationality, metacognition, or other related concepts. My sense is that you see intelligence, and thus rationality, as residing mostly in automatic, domain-specific processes, and associate agency with more controlled and general forms of reasoning that you think are more likely to undermine rationality than to enhance it.
Since getting this email last week I’ve been tossing it around in my head. I think I agree.
People know much more than they can say. Our verbal descriptions most closely match our behavior when we are new at a task, and know it only as a series of steps. With practice we no longer think about those steps — we automate them — so that we can perform them mindlessly while thinking about other things
The human ability to think has two main purposes: to allow us to learn new thinks (reorientation) and disrupt the execution of already automated tasks (disorientation). That is, thinking is a tool that should be used when our orientation is insufficient for the actions we have to perform. Normally, we rely on anxiety, or disorientation produced by orientation, to tell us when we need to calculate a new path or go back and reorient ourselves for a later time. Metacognition is similar to anxiety, except that it’s controlled by decision instead of orientation.
So why am I skeptical of rational agency, the idea that being human means having well-thought-out reasons for one’s actions? Because the tool of thought is just that, a tool. Decision is a tool used by persons in situations where they are unable or undesirious of trusting what they already know — it is not the essence of personhood.