Real pain is obviously very different from feigned pain-behavior, but we exhibit and express both in the same way. I cannot privately exhibit pain to myself in the way that I can publicly exhibit a broken tooth to others. In the case of others, the criteria for determining whether someone is in pain are the same for determining whether the pain is real or feigned; in my own case, there are no criteria at all.
When I understand a pattern in a series of numbers and say "Now I can go on!" why am I certain that this moment of inspiration will be followed by correctly writing out the series? It is odd to say that the relation between the moment of inspiration and
the feeling of certainty is causal or inductive. This certainty need not be justified by anything more than my proceeding to write the series correctly.
Our ways of talking about thought may tempt us to think that thought runs parallel to speech, as if thought were speech without words, so that when we speak we are reporting this inner monologue. But speaking is not just a matter of reporting the thoughts
within us. The very notion of having thoughts only makes sense regarding creatures that speak, judge, and question. We feel uncomfortable about the question of whether or not machines think, not because we think it unlikely that a machine could have an i
nner monologue, but because we are not even sure how to ascribe such an inner monologue to a machine.