Philosophy of Mind Part 11: Self-Knowledge

The Cartesian Picture

For Descartes, the mind was diaphanous to itself: every mental state was, by its nature, transparently available to its owner. To have a thought was to know one had it; to feel a pain was to be aware of feeling it. There was no gap between mental fact and mental knowledge. The mind was the one domain in which doubt could not get a foothold.

This picture had two strong components: infallibility (one cannot be wrong about one's own mental states) and self-intimation (one's mental states cannot fail to be known to one). Both have come under sustained pressure in the last 50 years.

Privileged Access

The weaker, defensible thesis: each person has a privileged epistemic position with respect to their own mental states — better in some way than anyone else's. The best contemporary account is asymmetry: I know my mental states by a different method (some form of introspection or rational self-attribution) than I know yours (behavioral inference). The methods need not be infallible; they just have to be different and, on average, better for the first-person case.

Even this weaker thesis has been challenged. Eric Schwitzgebel's Perplexities of Consciousness (2011) marshals evidence that introspection is dramatically unreliable even for ongoing phenomenal states — people give wildly inconsistent reports of whether they have visual imagery, of how their visual periphery looks, of what emotions they are currently feeling.

Transparency

A different idea, defended by Gareth Evans and Richard Moran. To find out whether you believe that p, you do not turn your gaze inward and inspect your belief; you turn your gaze outward and consider whether p is true. If you conclude p is true, you have thereby concluded you believe it. The question "do you believe p?" is, in normal cases, answered by addressing the question "is p true?"

Self-knowledge of belief is thus not introspection of an inner fact but the rational stance of endorsing a content. This explains why we know our beliefs without inner observation, while leaving room for cases (deep-seated prejudices we wouldn't endorse) where the transparent method comes apart from what we actually believe.

Nisbett & Wilson — Telling More Than We Can Know

Richard Nisbett and Timothy Wilson's 1977 paper "Telling More Than We Can Know: Verbal Reports on Mental Processes" was a watershed in psychology. Reviewing dozens of experiments, they argued that subjects' reports about why they made the choices they did were largely confabulated rationalizations, not accurate descriptions of underlying mental processes.

Confabulation Across the Brain

Neuroscientific evidence has dramatically reinforced the picture. Michael Gazzaniga's split-brain experiments showed the left hemisphere — the verbal "interpreter" — confidently constructing explanations for actions it had no causal knowledge of, when the right hemisphere had been the actual decider. Patients with anosognosia (denial of paralysis) construct elaborate rationalizations for not moving paralyzed limbs. Hypnotically induced behaviors generate sincere subjective reasons.

The picture is unsettling. Conscious self-attribution often appears to be a narrative-generating subsystem that runs after the actual decision, weaving a plausible story that the agent then sincerely believes.

Self-Deception

The classical paradox: how can someone simultaneously believe that p and believe that not-p, without recognizing the contradiction? Solutions include:

• Partitioned self (Davidson, Pears): The mind contains semi-independent subsystems that can hold contrary attitudes; conscious access privileges one over the other.
• Motivated belief-formation (Mele, Funkhouser): No literal contradictory beliefs, but biased attention and evidence-handling produce a comforting belief while suppressing engagement with disconfirming evidence.
• Avowal vs belief: One sincerely avows p while one's behavior reveals a contrary genuine belief.