[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Ball Lightning and "experimenter regress"



Original poster: Jim Lux <jimlux@xxxxxxxxxxxxx>

At 10:19 PM 6/22/2005, you wrote:
Original poster: William Beaty <billb@xxxxxxxxxx>


> Original poster: "Bob (R.A.) Jones" <a1accounting@xxxxxxxxxxxxx>
> In my opinion the biggest mystery of Ball Lightning is how many people claim
> to have seen it and how many people believe it exists.


This is unclear. Are you imagining that BL doesn't exist?



<snip>



A similar problem occurs whenever an experimenter makes a measurement, but
the expected value of the measurement is not known.  If you read the
voltmeter, yet have no "theory" that predicts an expected voltage, then
you can't be certain whether the meter is broken, or whether some other
mistake was made.  But then your goal is to create a new theory to explain
your voltage readings.  But then you need a theory before you can be
confident that your voltage readings are correct.  But how can you be
confident in any theory based on the questionable voltage readings you
took?  But you can only create new theories based on evidence such as
voltage readings!  An endless circle.

You can design another experiment to confirm that the voltmeter isn't broken. Or, develop sufficient theory (which may already exist) to reduce the probability that the voltmeter is broken to a very small number. (This is the "calibration" approach.)


However, there ARE cases where you can't do this.

In the case of BL, I think that there is sufficient theory out there to account for many of the observed behaviors. There are a few confounding factors though:
1) There's probably more than one phenomena described as BL. Are those unusually dark, hairy humans or a new species? Given that there are literally thousands of events where something described as BL has occurred, there's something going on, but exactly what is hard to know. Since there's more than one phenomena, there's likely to be more than one underlying cause: BL from sparking metal electrodes is going to be different than BL from soot.


2) Any one observer has probably only seen one, or maybe a few, instances. If you were to create a machine that could crank out some form of BL repeatably, then it's very hard to get a valid comparison. Each time you make a new BL instance, the observer is going to drift towards or away from thinking that it's just like what they saw. Even though "trained observers" have seen BL, they all view it in the context of their experience, and, almost inevitably, they weren't expecting to see it.

Consider eyewitness descriptions of suddenly occurring accidents (not those where "you see it coming" long in advance). They are notoriously unreliable, and exceedingly subject to modification by the interrogation protocol used to elicit the account. People have a hard time accurately remembering such things as the color of the car, if it's not sitting there in front of them.