I agree that these so-called "alternative medicines" have for the most part not been subjected to sufficient scientific tests, and that many of the studies that have been done are inadequate. I agree that herbs are just impure drugs.
However, it saddens me to see scientists refusing to consider the possibility that some of these therapies work, even if we don't (yet) know why. We scientists should be skeptical of claims but open to learning, or at least getting some new research ideas, from some of these alternative practices. Lumping all "alternative" things together and laughing at them does no one any good.
I know that one anecdote does not equal statistical evidence, but I also know what works in my life. I have a repetitive strain injury that has been treated using "standard" methods- high doses of NSAIDs, steroid shots, physical therapy, etc. These therapies helped, but the injury kept flaring up and I was facing a forced career change. Finally, I tried yoga (I may be the last person in Southern California to try it). One yoga class a week keeps me from relapsing, whereas a battery of physical therapy stretches done 3 times a day did not. Do I believe this is because yoga is tapping into my chakras or some other such thing? No. I don't know why yoga works and the physical therapy stretches didn't. Maybe it's the stress reducing effects of a weekly yoga practice- there is certainly a lot of solid scientific evidence about the negative effects of stress. Maybe the yogis stumbled onto some biomechanically optimal combination of stretches/poses. Too bad studies are unlikely to be done to try to understand what is going on. You can blame the alternative crowd for that, but frankly, if we scientists laugh at the very idea that practitioners of yoga, tai chi, or qi gong are experiencing real benefits and refuse to take seriously anyone who dares to try to investigate this, we bear some of the blame, too.
I am a firm believer in the need for good, well-designed scientific studies to establish the benefits of any treatment. Yoga has helped me, but that does not mean it is a good treatment for other people with similar repetitive strain injuries or asthma (I've also noticed a marked improvement in my asthma since I started doing yoga, to the point that I no longer take maintenance medications). I may have been experiencing a placebo effect (although I doubt it, since I went in fairly skeptical), or there may be something peculiar about my physiology or particular injury that means yoga does something for me that it won't do for other people. Only a randomized trial could sort these things out. I certainly would not advocate for people to disregard their doctor's advice and just go to yoga class. However, I think it is really a shame that careful studies are unlikely to be done, and as I say at the end of my comment on Derek's post, I blame both the alternative medicine practitioners for dismissing the need for studies and scientists who contribute to a climate in which anyone wanting to investigate "alternative medicine" risks being labeled a crackpot.
I have written before about how I am distressed by the fact that many people don't trust the federal scientists who are charged with determining what is and isn't safe. Maybe I shouldn't be surprised that people distrust and/or disregard science when things that they have found to be helpful and effective in their life are summarily dismissed by so many scientists. I'm not sure I'd want to listen to a group of people who told me I was gullible or naive for believing in my own experience, either. We scientists in the pharma and biotech industry need to do a much better job of explaining how what we do is different from what manufacturers of herbal and homeopathic remedies do, and why we don't think those remedies are good treatments in most cases.
And we need to be open to the possibility that some of the treatments and approaches labeled as "alternative" might actually be doing something. I won't believe something works without seeing the evidence, so I shouldn't assume something doesn't work without seeing some evidence. Skepticism should go both ways.