Thing is, this theory is perfectly testable – I foresee the need for 4 test groups of people with the same medical complaint – Group 1 will receive the homeopathically appropriate remedy for their condition, one group will receive double distilled water shaken/diluted further in the homeopathic manner but with no remedy included before the dilution, one group will receive double distilled water with no further shaking or dilution done to it, and the final group shall receive a homeopathic solution containing an inappropriate remedy that is meant to treat another condition entirely.
According to homeopathic theory, only the first group (A)should receive a therapeutic benefit if the dilution hypothesis is correct, and the third group (C) will receive no benefit but the others will if the agitation hypothesis is correct. If both dilution and agitation hypotheses are correct, then the first group A will have the strongest beneficial therapeutic response compared to groups B and D, who will both have a greater therapeutic benefit than group C. Of course, if neither dilution nor agitation change the water’s therapeutic properties, then all 4 groups will have the same level of therapeutic benefit – i.e. indistinguishable from the placebo effect.
All these different solutions would have to be administered to the test group in the proper double-blind manner – neither the patient nor the person giving the remedy knows what is in the particular bottle that the dose comes from. There are well understood study protocols for coded labelling of test doses that could be used for this.
Go on, homeopathy – do this and show us any significant variation in the therapeutic outcomes for the different groups. If you can, and if subsequent researchers can reproduce the results using the same protocols, THEN I will take the notion that water has memory and that shaking it produces quantum effects on molecule structure that affect body tissues at the nano level seriously. But until then, I just won’t.
h/t @shonias on Twitter