CCNet has a link to a Live Science article which says "Global Warming making North Atlantic less salty." This is followed by an article from New Scientist which says "Global warming is making North Atlantic more salty." I couldn't link to the New Scientist site as it is down for maintenance but the url is http://environment.newscientist.com/article.ns?id=dn12528&feedId=online-news_rss
which you can try when it comes back on line, from the August 23rd issue.
This is just one more example of why the general public distrusts science. A Joe lunchbox will see headlines like this and wonder something like "Why can't these scientists get it right?" Indeed. But this just highlights the fact that studies can be made to show just about anything. In reality they really can't prove anything. They're not designed to. They deal in probabilities. Their value to science lies in a study's ability to narrow down possible causes to a few at which time science then conducts a test to prove (or disprove) a causal connection. Without that test or experiment, there is no certainty or proof of causality, only a maybe.
Today, actual experiments are being ignored and statistical correlations and associations are being treated as if they were causal connections by use of the word 'link.' We constantly hear things like "cause A is linked to disease B" and so on. Yet the dictionary definition of link is connection or joining. But correlations are not connections or joinings. The public hears the word 'link' and wrongly assumes a causal connection.
The growing indifference to using experiments or tests to prove causal relationships, or to intellectual precision by calling associations 'links' is further evidence that "Government encouragement does not order men to believe that the false is true, it merely makes them indifferent to the issue of truth or falsehood." (Ayn Rand, Establishing of an Establishment.) The solution is to dismantle the establishment.
[CCNet is a scholarly electronic network edited by Benny Peiser. To subscribe,
send an e-mail to firstname.lastname@example.org ("subscribe cambridge-conference").] It's free.
Update: The link to the New Scientist article is here.