Flim-Flam and the importance of repeatability

Amba sent me this link with the description "Interesting stuff." It's a critique of one chapter in James Randi's book Flim-Flam. Here are my thoughts.

First, to the credit of the author (Michael Prescott), he links to another version of the article with a brief response from Randi, as well as links to two relavent pieces on Randi's website. Prescott's counter-rebuttals to two of those points, about Dr. Hebard and the tape recording, end up looking rather pathetic, in essence saying, "I never said the accusations were substantiated in the first places." On the issue of the tape recorded conversation where Randi allegedly admitted to being wrong on every single one of a list of 24 points, Prescott quotes Puthoff giving a rather, shall we say, grandstanding resonse:

Of course, having got caught, Randi would have to call me a liar, and count on the fact that I would be unlikely after all these years to put my hands on the tape. You can quote me in saying that I say that Randi is a liar when he calls me a liar. My profession is as a scientist dedicated to reality and truth, his is as a charlatan dedicated to misdirection and tomfoolery to gain the moment. Let the audience figure out who is more likely to be lying!

Well, putting my best figuring ability to work, I figure it's unlikely anyone would make such a recording only to lose it. This also makes Puthoff's claim of having a signed affadavit contradicting Randi on another issue suspect.

Another miscellaneous note: Prescott notes, with an unfavorable tone, that Randi, "reports that their experiments were conducted in a chaotic atmosphere conducive to cheating." Interestingly, they chose to title one of their own accounts of the experiments "Is Chaos Necessary?"

The above points I make for the sake of a more overall look at the article, and may not be of interest to most readers of this blog. There is a more general point to be made here, however: First, I wish Randi's critique of Geller's stunts was not not based on second-hand information ("Chinese Whispers," as Prescott calls them). I also wish Geller had simply been willing to do a repeat performance with a different group of scientists willing to fix alleged errors. To quote from an attack on Randi's Million Dollar Challenge:

A leading Fellow of CSICOP, Ray Hyman, has pointed out, this "prize" cannot be taken seriously from a scientific point of view: "Scientists don't settle issues with a single test, so even if someone does win a big cash prize in a demonstration, this isn't going to convince anyone. Proof in science happens through replication, not through single experiments."

This statement only bolster's Randi's case when he's asking psychics to repeat feats they supposedly performed in other circumstances. And repetition really is important, not just because some scientists say so. The universe is a complex place, and it can be nearly impossible to know the reason for an experimental result just by after the fact analysis. This applies to the most mundane of fields, though it arguably goes double when tricks are suspected, as tricks tend to rely on seemingly irrelevant details that observers quickly forget. Repeating an experiment provides a much easier way to find sources of error. Without such a repeat performance, after the fact guess work is all the analysis anyone can do.

Though Geller never, to my knowledge, repeated the exact experiments run by Puthoff and Targ, it is worth noting that when other SRI researchers tried similar experiments with him, he failed miserably. His excuse? "Negative vibrations."