Thought experiments occupy an uncomfortable position in accounts of how knowledge is produced, since they appear to deliver substantive conclusions about the world without any new observation of it. Galileo's argument that two bodies of different weight must fall at the same rate — the heavier and the lighter tied together falling both faster and slower on the Aristotelian view — is the standard example, and it does seem to establish something. The deflationary reading holds that such arguments merely make explicit what was already contained in the premises, rearranging commitments rather than adding to them. This is probably right and rather less deflating than it sounds, given how much of any theory is implicit and how rarely its consequences are apparent to those holding it. Making a contradiction visible is not a small achievement, and it is not the same as discovering a new fact.