Infer from similarity to oneself.
Shared embodiment and behavior can support attribution, but similarity is graded and theory-laden.
Side 116
Other Minds
How can one know that another entity has experiences, beliefs or intentions when only one's own mental life is directly accessible?
observation→inference→model→attribution→uncertainty
04lenses
16working concepts
V0content
SS-1.0standard
Other minds are not observed in the same way as behavior.
The problem begins with the gap between observable action and attributed inner state.
Mental states can explain organized behavior.
The inference is stronger when rival non-mental explanations fit less well.
Use models of beliefs, goals and perception.
Attribution depends on assumptions about what the other agent knows and wants.
Some views reject a large inferential gap.
Expression and embodied interaction may present mindedness more directly than abstract analogy suggests.
Different evidence types support different claims.
Behavioral flexibility, learning, communication and embodiment each constrain attribution without settling it alone.
Look for flexible goal-sensitive action.
Complex behavior can indicate internal organization while still permitting multiple mechanisms.
Track adaptation across experience.
Learning narrows explanations based on fixed stimulus-response rules.
Test shared reference and responsiveness.
Communication can reveal perspective-taking, but apparent language competence can be produced by different architectures.
Consider sensorimotor organization.
Shared biological structure can support analogy, yet mindedness may not require identical embodiment.
Animals and AI stress-test human-centered criteria.
Cases unlike adult humans reveal which evidence actually matters and which assumptions were merely familiar.
Species differ in sensory worlds and action repertoires.
Absence of human-like behavior is weak evidence when the organism solves different ecological problems.
Mind attribution precedes mature language.
Developmental cases show that verbal report cannot be a universal criterion.
Linguistic performance complicates attribution.
Behavioral fluency is evidence about functional capacity, not direct evidence of phenomenal experience.
Groups can display organized agency.
Whether institutions literally have minds depends on the theory of mind being used.
Uncertainty should be preserved rather than hidden.
No single behavioral test establishes consciousness or mentality across every possible system.
Multiple mechanisms can fit the same behavior.
Successful prediction does not uniquely identify inner architecture.
Human categories can overfit unfamiliar systems.
Similarity in language or appearance can bias attribution.
Difference can also be overinterpreted.
Rejecting mindedness merely because a system is unlike humans is the mirror error.
Uncertainty can still matter morally.
When stakes are high, decisions may depend on the cost of false attribution versus false denial.
Mind attribution is an inference under uncertainty. The aim is not to eliminate uncertainty but to make the evidence, assumptions and competing explanations explicit.