I'll admit up front that the HIV/AIDS debate is beyond my present competence. I can't judge one way or another without a lot of additional research, and I have a lot of caution flags when wading into that kind of territory. Still, those articles and many of the intelligent commenter responses are a good window into some of the central dilemmas about science in public policy - and how we deal with it as individuals and as a society.

Having referred y'all to Dean's posts and guest blogs from researchers in this field, I'm going to talk about that window - and about how we can apply intelligent skepticism when stuff is beyond our personal depth. Comments welcome.

I'll start with some things to watch out for:

• The majority rules fallacy. Science isn't about majority rule. Science is about being more right (or, as some have put it, "more subtly wrong"). There are always guys out there like Tesla, and there's also a long history of the "moonbats" ending up with the better theory, hence Kuhn's work on the structure of scientific revolutions that gave us the overused word "paradigm." Just because lots of scientists believe something doesn't mean it's true.

Dean has links to an excellent real-life demonstration of this truth. Pellagra was a killer in the U.S. south, and getting a handle on it was tough. He explains how that went, and clicking the multimedia presentation he links is a great way to really understand the scientific method in about 15 minutes while following an interesting mystery.

Having said that, weight of opinion has some value. Hence the need to add a caution about:

• The "he said/ she said" effect. The media like to have 2 sides to a story. Controversy sells, even though there may be a strong scientific consensus around one side. Just because you see 2 opposing points of view in print, therefore, doesn't mean they have equal weight. Science isn't perfect, and scientists and their organizations are prone to all the same public choice theory problems and orthodoxies as other human organizations. But the collaborative filtering of hundreds of scientists doing and repeating hundreds of experiments is an important part of the scientific method, and one disregards its weight at one's peril.

• The troll pattern effect. Many great scientists lack interpersonal skills - but rigor when discussing their subjects is something no good scientist will compromise. All the usual troll behaviours of misrepresenting opposing views, changing the subject when confronted by inconvenient facts, "hand waving" dismissals without links or back-ups, hypotheses resting on nothing or on emotion, unwillingness to concede even a single point, persecution complexes and conspiracy theories, etc. happen in scientific debates too. The more of them you see, the less credibility you should lend to that person's arguments.

• The troll is not the theory. If you do find a scientific troll, all you've done is dismissed the troll's argument and put the issue to bed temporarily. A better presentation of that same general point may come along later, and it may be correct.

Avoiding obvious errors is the first step to wisdom, and remembering these rules will help us avoid jumping to foolish conclusions. How do we go on from there to make positive judgments about scientific public policy issues?

• Follow the Basic Rules: Disprovable. Measurable. Controlled. Repeatable. If someone can point out weaknesses in any of these areas, it may be game over quickly.

1. Scientific hypotheses must be disprovable. If you can't disprove it, it's not scientific (which is why creationism isn't science - you can't disprove G-d's existence or reported actions).
   
   
2. On a related note, they must be linked to something measureable - otherwise, how can you disprove them? Advertising that says "X will make you more beautiful" isn't scientific, but "having people drink 5 mugs of X made 70% of them improve their beauty ratings for the opposite sex" would pass the test.
   
   
3. Experiments must be controlled, to see if there's a difference from a random group. Otherwise, all you may be measuring is something caused by your selections or by the outside environment. "Russians (unsaid: in Moscow) who took our pills had 40% less radiation damage than the rest of our Russian test goup (unsaid: from Chernobyl)" is clearly cheating.
   
   
4. Scientific results must be repeatable; people doing the same experiment must get the same results, so we know it wasn't a fluke.

Other yardsticks you can use to evaluate scientific controversies include:

• Question quality + verification. A good argument should be able to raise key questions about existing scientific beliefs that clearly explain existing beliefs, what we should logically expect to find, and what's different. You don't know that they're right yet, but if that chain stands up to scrutiny you can strongly suspect that something is wrong with existing theories.

But verify. Get confirmation that their explanation of existing beliefs is correct (many creationists either don't understand evolution, for example, or they deliberately misrepresent it). Do you see a logical connection between those beliefs and what they say we should expect to find? Do they use the words of the present theory's supporters, or reference their works so you know they're being fair? Can their notes re: what's different be explained by something else?

• Elimination. Use the "stuff to avoid" as your guide, in order to improve your odds. If you find one side not playing fair or breaking many of the rules above, you can either eliminate the challenge until a better example comes along or increase your suspicion that something's wrong with existing views.

When looking at existing views, pay special attention to major statements and documents that are often referenced; if they're playing fast and loose, that's a red flag. When looking to 'handicap' dissenting views, pay special attention to patterns of similar kinds of shortcomings demonstrated among multiple opponents.

• Better Alternative Explanation. Science can discover that something's wrong with existing theories, without knowing what's right yet, and be perfectly valid. If an opposing theory is put forward, however, and if it holds up to scrutiny and explains a broader range of results (including some puzzlers the old theory didn't explain), that's a good indication they may be on to something.

Each of these approaches can help move your "suspicion meter" when confronted by conflicting science claims or by pseudoscience. They're not perfect, but they're often helpful... and understanding them better helps you understand what science is really all about. It isn't perfect, and it doesn't have the answers to every question - but it is the foundation of our society and its achievements.

Which is why shortcuts will only get us so far, and why we have to prepare ourselves as a society by insisting on strong science education in high schools and in universities - even in the liberal arts programs. It may not be rational for people to become experts on HIV virology, but someone who doesn't understand the basics of a field is utterly lost from the get-go. They will be easy prey for charlatans - and they will lose the argument even when they're right.

That's why we have to work on giving people an adequate set of basic knowledge on which to rest important discussions, including knowledge of science and the scientific method itself. A culture that loses that becomes a Dark-Age culture of serfs, Morlocks, or both.