An NT (neurologically typical, non-autistic) woman commented that, although a major diffrence between "normals" and autistics is that "normals" are said to be able to "read" non-verbal signals, even "normal" people often "get it wrong." I responded:
Maybe one tentative conclusion to draw from the above is that quite often "being wrong" doesn't make a big difference. Sometimes it does, obviously, and can lead to big problems in interpersonal relationships. But if NTs are "wrong" about the way they interpret casual conversations (presumably because they tend to interpret from their own POV, which may not be shared by the other person, and perhaps also on the basis of more-or-less subconscious preferences), it may often be a case of "no harm done."
Being "wrong" without consequences may tend to reinforce behavior, so NTs -- not realizing how often they are "wrong" -- may gain unrealistic confidence in their own interpretive powers. (This would cause major problems for their interaction with autistics, BTW, because the more sure of themselves the NTs are, the harder it will be to persuade them to try a more open-minded approach.)
Maybe what's important for NTs is the (more or less unconscious) act of interpreting social input. Whether they are right or wrong may be less important for them than their sense that they are participating. When they meet up with an autistic whose forms of communicating do not allow for the full experience of signal-sending and signal-receiving, it may be the lack of mutual participation that is most troubling for the NT person, rather than a feeling/realization that s/he is being "wrong" in interpreting what the autistic person is communicating.
Just playing with ideas here, folks.