The distinction between fact and belief is actually a complicated one and whole books have been written about it, so trying to pull the different levels of confusion apart would be far beyond the scope of this discussion I believe. Nevertheless I agree with tarvos: Most people don't really know how science works. Even the scientists themselves are usually not aware of all the politics involved, because they have never studied Philosophy of Science. They are usually also completely oblivious of the real implications of their findings. In some cases they might be aware, but then exploit the naiveté of journalists and funding bodies to promote their work beyond its actual merit. And then some "sciences" aren't scientific enough in their methods strictly speaking. In the social sciences for example there are lots of published studies that would make any proper statistician cry. That's why I generally don't take quantitative studies seriously unless the stats were done by someone with a proper science background (almost never the case) and prefer qualitative studies in general. Qualitative studies in the social sciences are essentially based on anecdotal evidence, i.e. single case studies and those tend to hold more truth than most quantitative studies, because many of them simply use the wrong kind of statistical methods to analyse the data. This isn't only the case in the humanities though, one can poke holes into many studies in the sciences as well because of too small sample sizes, confirmation bias and so on. Medicine is the worst offender usually, because the studies are paid by the pharma industry and have an extreme financial interest in positive results. Completely agreed, PM! And then there is of course the big problem that 99% effective or 99.999% without serious side-effects usually means that I'm the 0.001% case, because I tend to be super unlucky in all medical things. It doesn't really matter whether 99,999 other people came out of a treatment alright, when you're the one person who now only has 20% kidney function left. Statistical approaches have their limitations, especially for people who fall outside the norm, a big flaw in medical research where the consequences usually matter much more than in other fields of research. And by the way, the preference for positive results is actually a problem in all of science in general, negative results are usually unpublishable.
tarvos wrote:(that's why Lamarckian genetics isn't used any more, but other genetic models are).
Actually, Lamarckian genetics is coming into fashion again without his name being invoked much due to the usual "politics": Epigenetics is huge right now and that's pretty Lamarckian. Of course Lamarck wasn't entirely right, but one can't say that he was completely wrong either, because development indeed influences the expression of genes. I'm sure Lamarck will get a place of honour again once Epigenetics makes its way into school biology books, although that will obviously still take a while. Another problem in the sciences: Some terms simply get a bad reputation for a while, because they don't fit current fashions. And scientists who don't follow current fashions are pretty much unemployable. Certain terms or findings of certain scientists come in and out of fashion in waves and usually a whole generation of scientists prefers one theory over another. Latour writes a lot about these movements in science. And Latour himself is of course hugely fashionable in the humanities right now precisely because he pulls the sciences down to the same level as the humanities: unreliable, dependent on fashions, heavy on interpretation and so on.
Basically we're on unstable ground wherever it comes to "knowledge". And when you keep that in mind, the whole notion of "scientific proof" becomes an unattainable ideal. In the end, knowing and believing is generally much closer together conceptually than assumed by the general population AND the scientists.