
Trials and Errors: Why Science Is Failing Us 474
Lanxon writes "An in-depth feature in Wired explores the reason science may be failing us. Quoting: 'For too long, we've pretended that the old problem of causality can be cured by our shiny new knowledge. If only we devote more resources to research or dissect the system at a more fundamental level or search for ever more subtle correlations, we can discover how it all works. But a cause is not a fact, and it never will be; the things we can see will always be bracketed by what we cannot. And this is why, even when we know everything about everything, we'll still be telling stories about why it happened. It's mystery all the way down.'"
Science isn't a goal (Score:5, Interesting)
It's a direction.
Re:Everyone a specialist now (Score:5, Interesting)
I agree.
I think we need to start focusing on systems theory. Many large systems share some very similar characteristics. We need people who are big picture people, who can see the forest for the trees. Of course, without knowing about the trees, a forest is something of a mystery. We need both kinds of people. But the usefulness of pure reductionism is at its end, and we need to recognize that and start taking a different approach to understanding.
Re:Everyone a specialist now (Score:5, Interesting)
"Many large systems share some very similar characteristics...We need people who are big picture people, who can see the forest for the trees."
Except that everyone who gets large systems dropped out of the current, fucked-up system long before being awarded a research post for their willingness to play along.
Re:Then we must live forever (Score:4, Interesting)
Your clone wouldn't, but you'd be dead. I doubt having two copies of the same brain suddenly make your consciousness share both of them.
When the original dies, you die. An undistinguishable copy of you lives on.
Re:Everyone a specialist now (Score:5, Interesting)
We need people who are big picture people, who can see the forest for the trees. Of course, without knowing about the trees, a forest is something of a mystery. We need both kinds of people.
I think it's a mistake to think that these should be two different groups of people. There are a lot of "forest" people who don't actually know anything at all about trees, and whatever they think they know about forests will be complete nonsense as a result. You see this a lot on Slashdot, actually; it seems to be a common failing among computer scientists to think that just because you can write code to describe a system, in some fashion, that means you actually understand the system. Certainly scientists in a lot of fields tend to overspecialize, but in interdisciplinary fields such as bioinformatics, you just have to start with some of the tree knowledge, or you won't be able to say anything meaningful about the forest at all.
And yes, this means spending a lot of years in school studying many different and not-obviously-related subjects, and no, that blog post you read last week doesn't count.
Re:Everyone a specialist now (Score:4, Interesting)
Realising you need the partnership is half the problem. :) Mathematicians and computer scientists didn't show up and say "Look, guys, you need to stop looking at this thing one gene at a time." The field was essentially founded as a result of Fred Sanger's early work with whole genomes, and he was a biochemist to the bone.
At any rate, the distinction doesn't matter; the point is that the problem of scaling up and looking at the big picture was resolved in the case of the biological sciences, and the view continues to get broader through approaches like environmental sequencing and metagenomics. The problem described in the article is exactly a case of an old-school, low-throughput mindset and insufficient concern for other variables. Reductionism can work very well when you don't accidentally leave things out! The trick lies in only reducing the system once you have good reason to believe that you've ruled out all the other possibilities.
Re:Everyone a specialist now (Score:4, Interesting)
The predictive outcome is you mention is otherwise known as a scientific theory, and is pretty much what science is all about. I don't care much whether a theory is "reductionist" or "systems", as long as it's a good theory (it works!), it's valuable. I do agree that many science teams could use an outsider systems guy to try an see the big picture better by not being absorbed in the minutiae.
Re:Then we must live forever (Score:5, Interesting)
Difficult, sure, but saying 'impossible' is pretty much walking around with "I have no idea what entropy actually means" tattooed on your forehead. The human body is in no way a closed system, and the second law of thermodynamics says nothing about the change of its entropy over time as long as it has an energy input and a universe-sized heat sink to dump excess entropy into.
Cancer just means that evolution is hack piled upon hack until it stumbles onto something, so it does much better than human engineers at designing really complex interacting systems without very much abstraction or modularization, but much worse at discovering things which you'd never, ever stumble onto without conceptual understanding, like Reed-Solomon codes. If it had, then it could make the mutation rate exponentially low for only a linear increase of complexity and energy requirements for manipulating genetic material, and cancer would be worth worrying about roughly as much as brute force attacks against AES-256.
Of course, re-engineering such a fundamental, low-level feature of an organism might very well be harder than just designing a new one from scratch, but 'impossible' doesn't pass the giggle test. There's nothing anywhere in the laws of physics to say such a thing is any less possible than the mutation-prone organisms we already do have.
The hacker koan of the randomized neural net (Score:2, Interesting)
Similar point in a different form:
Hacker koan: Uncarved block [wikipedia.org]
Re:Then we must live forever (Score:4, Interesting)
Why do these organisms have longer lives than us?
Fewer moving parts?