Seems to me, landis, like you’re going about this all wrong–I think this abstract contemplation on the logic of your experimentations is a red herring; keep it simple–the way you described your plan for experimentation would do just fine:
But since this is a philosophy forum, why don’t we address your concerns over the logic of this whole thing.
I had an interesting discussion about the logic of conditionals recently: viewtopic.php?f=1&t=185298.partial
Essentially, the only reason why the above is inferentially invalid is because eating nightshades may not be the only thing that gives you migraines. Hangovers might do it; getting bashed over the head with a baseball bat might do it.
The reason why denying the antecedent seems to work for you is that all this stuff is implicitly assumed–you’re not just thinking that abstaining from nightshades will eliminate your migraines but also abstaining from a night of heavy drinking, or getting bashed over the head by a baseball bat, plus anything else that might give you migraines.
Yeah, what’s wrong with that?
Is this what Popper did? I don’t get it.
Right, which is another way of explaining why your example above about denying the antecedent seems to work for you: just because you get migraines doesn’t mean you must be eating nightshades–as I said above, your migraines could come from a whole swack of different things.
I think you’re getting ahead of yourself here. Let’s not assume tolerance comes into play at first. If you find that abstaining from nightshades eliminates your migraines, problem solved. If not, then try the test again controlling for tolerance.
Modus Tollens wouldn’t falsify anything.
If I drink poison, then I’ll die.
I’m not dying.
Therefore, I didn’t drink poison.
This does not falsify the claim that drinking poison causes death.
I don’t think there’s any logical way of putting it in terms of affirming the consequent. Like I said, affirming the consequent only seems to work because you’re bringing in a bunch of implicit assumptions.