I believe it was Ryle who, in defending his outright silly view that mental properties are merely properties of our physical bodies, i.e. our behavior, insisted that their must be some sort of motion that we do, even if undetectable, to correspond to our innermost thoughts. It's true that some people do this, but most people don't move their lips when thinking. This was rightly seen as special pleading.
NASA scientists have now shown that Ryle was right on one thing. When we're about to talk, some sort of behavior in our throat preceds our mouth's uttering of words. These motions can now be analyzed and given content. Computers can now tell what you're saying subvocally even if you don't move your lips or face.
Now this doesn't really rescue Ryle. His version of behaviorism is demonstrably false, which is shown by a number of cases:
1. The in-principle possible perfect actor spends an entire life acting out the behavior that accompanies certain mental states without ever having them.
2. The in-principle possible muscle relaxant that doctors believe to be a perfect pain-killer. They never see the effects of the pain caused by the operation, because every measurable muscle in the body is temporarily frozen somehow, preserving life and feeling. Then the drug's amnesiac properties kick in, and the person forgets all the pain caused by the operation, so no one ever finds out through behavior that there was any pain. Yet there was.
So Ryle was wrong that pain or other mental states can be viewed as simply behavior or dispositions to behave in certain ways. However, the irony is that that the most ridiculed element of his defense of such a ridiculous view has now at least partially been vindictated.