[edkennedy]

Interestingly this all started in 2017 when a PhD student started aggressively pursuing water companies to find out which ones used dowsing: https://www.theguardian.com/business/2017/nov/21/uk-water-fi...

It seems she did not follow up on the claims of accuracy, or investigate the validity of the practice at all. It was simply immediately discredited as witchcraft and called a waste of public money. I guess this is the scientist vs engineer dilemma at work. The fact that 10 of the 17 water companies were using dowsing 5 years ago before a public shaming shows that the practice could have practical use.

[guitarbill]

> The fact that 10 of the 17 water companies were using dowsing 5 years ago before a public shaming shows that the practice could have practical use.

Err, no it doesn't. By this argument, thousands of doctors having used bloodletting or trepanning also "shows that the practice could have practical use".

> It seems she did not follow up on the claims of accuracy, or investigate the validity of the practice at all.

She's the one who cited studies where dowsing performs no better than chance, while the water companies published no data on the matter.

[LawTalkingGuy]

fwiw, both bloodletting and trepanning have legitimate modern forms.

In this case, where many for-profit companies were practicing dowsing it makes sense to ask if it might have worked. Probably, I agree, not because it fundamentally works, but as others in this thread have suggested - because practitioners unconsciously recognize the external signs of leaking pipes. This could match the the studies - actual randomized dowsing wouldn't work but standardized dowsing of the usual failing pipes could be easy.