On today’s post, John Dorsch, a postdoctoral researcher at the Czech Academy of Sciences, introduces his new book, Origins and Future of Self-Knowledge. He explores how we come to understand and critically evaluate our own beliefs.
In Origins and Future of Self-Knowledge, I investigate the bodily, social and cognitive conditions that make this capacity possible. The book examines how we come to understand our own beliefs as beliefs: states that can be mistaken, require justification, and ought to answer to how things actually are.
 |
| John Dorsch |
How do we know that we know? We routinely distinguish between merely believing something and having good reasons to be confident. Yet the capacity to make these distinctions is remarkable. How did creatures like us acquire it?
In Origins and Future of Self-Knowledge, I investigate the bodily, social and cognitive conditions that make this capacity possible. The book examines how we come to understand our own beliefs as beliefs: states that can be mistaken, require justification, and ought to answer to how things actually are.
 |
| Book Cover |
Consider an ordinary exchange. Someone asks where the post office is. You point across the street. They ask, “Are you sure?” Here you are being asked to assess your own judgment and possibly take responsibility for its reliability. Hence this everyday interaction reveals a sophisticated achievement: recognising that what you believe might differ from what is true.
One starting point is what psychologists and philosophers call “noetic feelings”: experiences of confidence, doubt, familiarity and knowing. Think of having a name on the tip of your tongue. Although you cannot retrieve it, you feel that further effort might succeed. Such feelings help organise our thinking, directing when we persist, seek assistance, reconsider, or stop.
Explaining these feelings as internal feedback signals leaves important questions unanswered. What distinguishes them from the mechanisms through which any adaptive system corrects its behaviour? How could a feeling of confidence explain our ability to recognise that a belief needs reasons? Successful self-regulation alone does not explain accountability to truth.
I argue that answering these questions requires looking beyond the brain in isolation. Some noetic feelings depend on intricate relationships between brain and body. Bodily processes help make uncertainty or familiarity emotionally felt and capable of guiding thought. These experiences contribute to the control of attention, memory, and cognitive effort.
Their significance also depends on our relationships with others. Expressions of confidence and doubt become available for interpretation, challenge, and correction. Through practices of asking questions, offering reasons, and acknowledging mistakes, we learn standards for evaluating beliefs. Self-knowledge emerges through the interaction of embodied feelings and socially sustained epistemic expectations.
This account has implications for artificial intelligence. Technologies that enter our practices of questioning, checking, and justifying may reshape the conditions supporting our self-knowledge. Understanding those conditions gives us a basis for examining how AI might strengthen or undermine our capacity to evaluate our beliefs, our epistemic agency.
The account also opens a further possibility. Its explanation of how we become self-knowing agents is functional and structural, independent of the particular material of which we are composed. The relevant capacities could therefore, in principle, be realised in artificial systems.
The final chapter explores this possibility: a self-knowing machine. Could an artificial agent develop the capacity to understand its own beliefs as answerable to reasons and truth? Tracing the origins of human self-knowledge provides a framework for asking what would make such a machine possible, and what participating in our shared epistemic practices might require.
One starting point is what psychologists and philosophers call “noetic feelings”: experiences of confidence, doubt, familiarity and knowing. Think of having a name on the tip of your tongue. Although you cannot retrieve it, you feel that further effort might succeed. Such feelings help organise our thinking, directing when we persist, seek assistance, reconsider, or stop.
Explaining these feelings as internal feedback signals leaves important questions unanswered. What distinguishes them from the mechanisms through which any adaptive system corrects its behaviour? How could a feeling of confidence explain our ability to recognise that a belief needs reasons? Successful self-regulation alone does not explain accountability to truth.
I argue that answering these questions requires looking beyond the brain in isolation. Some noetic feelings depend on intricate relationships between brain and body. Bodily processes help make uncertainty or familiarity emotionally felt and capable of guiding thought. These experiences contribute to the control of attention, memory, and cognitive effort.
Their significance also depends on our relationships with others. Expressions of confidence and doubt become available for interpretation, challenge, and correction. Through practices of asking questions, offering reasons, and acknowledging mistakes, we learn standards for evaluating beliefs. Self-knowledge emerges through the interaction of embodied feelings and socially sustained epistemic expectations.
This account has implications for artificial intelligence. Technologies that enter our practices of questioning, checking, and justifying may reshape the conditions supporting our self-knowledge. Understanding those conditions gives us a basis for examining how AI might strengthen or undermine our capacity to evaluate our beliefs, our epistemic agency.
The account also opens a further possibility. Its explanation of how we become self-knowing agents is functional and structural, independent of the particular material of which we are composed. The relevant capacities could therefore, in principle, be realised in artificial systems.
The final chapter explores this possibility: a self-knowing machine. Could an artificial agent develop the capacity to understand its own beliefs as answerable to reasons and truth? Tracing the origins of human self-knowledge provides a framework for asking what would make such a machine possible, and what participating in our shared epistemic practices might require.