QuoteIn the context of a table-top experiment, the way one would distinguish error from not-error is by doing two otherwise identical experiments, controlling only for CCS, with one run intended to exhibit it, and the other intended not to exhibit it, and then looking at the difference and spread in the measured heat. This is the empirical method. You cannot explain away empirical results using only models and a speculative source of error.
That is true, but it may be as difficult as to find controls for the FPE results. D versus H is in no way a convincing control - given the physical properties of these two are so different - and RAE could just as easy be specific to certain metals and preparations as NAE.
The irony here continues to strike me.
But - there is a better way to attack Kirk's ideas - one that it surprises me has not been done. Which is to look in detail at the tolerance analysis in the context of specific experiments and prove him wrong, at least for those experiments.