X-Message-Number: 30066 From: Date: Mon, 26 Nov 2007 12:14:39 EST Subject: correction & addendum Correction: In my previous post I noted that D.S. had said that he knew of no secure basis for extrapolation other than existing data, and I said that I know of no basis for secure extrapolation including existing data. That was sloppy. By definition, extrapolation means projection of existing trends or data, so there is no other kind. My conclusion stands, of course--long term extrapolation is rarely reliable, unless you count such phenomena as the sun rising. Addendum: D.S. acknowledged that referreed journals contain lots of errors, but fewer than other sources. This is probably true in general, but misses the point. The point is that, in the minds of many people, the journals enjoy such prestige that published conclusions are accepted uncritically, even if the conclusion is not only wrong but egregiously wrong. The latter was the case in the Science piece I had mentioned as the possible source of D.S.'s statement about long term gains in life expectancy of 3 months per year. For a trend to be regarded as relatively reliable, it needs three things. One is regularity on the basis of numerous data, but that alone is not enough. Anyone can "mine the data" to find all sorts of strange coincidences. The second requirement is consistency with the larger body of knowledge. (e.g., Rhine's "evidence" for psychokinesis did not comport with well established physical law.) The third requirement is a rationale, some way of tying the phenomenon to natural causes, quantitatively and in detail--again something that Rhine lacked and that the "3 months" nonsense also lacks. Robert Ettinger **************************************Check out AOL's list of 2007's hottest products. (http://money.aol.com/special/hot-products-2007?NCID=aoltop00030000000001) Content-Type: text/html; charset="US-ASCII" [ AUTOMATICALLY SKIPPING HTML ENCODING! ] Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=30066