"andy-k" wrote in
news:SPd9k.87676$zs1.22385@newsfe28.ams2:

See previous short post. The testable consequences need not be behavior
(with respect to the general argument that predications of any property
must be cognitive). But there must be testable consequences of some kind,
if we are to be able to distinguish true predications from false ones. That
is the most convincing argument I can offer.

But given the conceptual ties and known empirical relationships between
subjective states (our own, and those reported by others) and behavior,
some sort of behavioral consequence is surely what we would seek (for any
subjective state denotable by some term in the public language).

Suppose a child says, "I feel cold." How might Mom respond? She might first
check the thermostat (and find that the room is a comfy 75 degrees). She
might feel the child's forehead, and decide she feels normal (she might
even take her temp with a thermometer). Mom might then check for goosebumps
or shivering. Seeing none, she may then begin to wonder whether the child
really feels cold. To test, she might offer the child a sweater or cup of
hot chocolate, both of which the child refuses. Then Mom says, "Wanna sit
in my lap a while and I'll read you a story?" The child scrambles up. After
a minute Mom asks, "Still cold?" The child says, "I'm fine now."

Did the child earlier experience the subjective state of feeling cold, or
did she actually experience a different subjective state