Skeptics: How do you know what to "believe"?

It is not possible for anyone to understand every policy-important aspect of scientific knowledge at the level of detail necessary to accept that knowledge as valid, or to defend it against the evil anti-science denialists. So what is a skeptic to do?

[a timely repost]

Real science happens at the very edge of knowledge. If you go do real science for a while ... a few weeks ... then go back to science geek land where you normally live (and I know you live there because you are reading this blog) and read about the same topic you were exploring in real science land, you will see two different worlds. When a scientist doing actual scientific research is at work, and she gazes into a microscope, or a matrix of data from the latest sample run, or the profile just exposed by an ongoing excavation, she may be looking into an unordered mess of great potential or deep disappointment that does not yet fit into what she knows. If she is lucky, it never will fit. It will reshape what she already knows in a slightly uncomfortable way, or tie up a lose end while slicing open another, or something. Or, it just won't work, and it will turn out that the sample is messed up or the trench was dug in the wrong place.

Most importantly, for her to understand the meaning of the newly observed data, she will have to have a great deal of expertise. Some of this expertise will be based on her own education, some from her experience with the particular problem she is working on, some garnered from colleagues with whom she consults, some from the generally available literature, some from in-house reports and other documents.

For a given cycle of data collection and analysis, which might incrementally change what is thought about a particular scientific problem, there is a story. Go ask the scientist "So, what happened at work this week?" and you might get that story (depending on the week) but if you are not already expert in that specific area, it might take a very long time for her to explain everything you need to know to really get it. And that can be the case for a tiny bit of some aspect of some molecule, or one aspect of the paleobiogeography of some snail, or whatever.

(This, by the way, is why one so often gets a flip answer from a scientist when you ask them what they are up to ... most researchers develop a handful of standard responses to address people's inquiries which ultimately avoid actually answering the question.)

But from what we often hear from those passers by, those with some interest in science, or those with a politically motivated denialist agenda, is so far from this kind of understanding that it can be very frustrating. It sounds like this ... someone brings their car in to have it fixed and they are talking to the person behind the counter at the mechanics.

"My car is not running right. I think the elves that make it run have tummy aches."

And then the mechanic tries to explain that there are no elves in the car, but the person insists.

"You have not seen that there are no elves in the car. And besides when you open the hood the elves become invisible. Anyway, the elves have tummy aches. Fix the tummy aches." And so on.

That is what many people who are not scientists sound like when they are talking about science. The discordance between what is on one hand being said and sometimes insisted or demanded or required, vs. how things actually work is vast. Cars do not have living organisms inside them (usually). So insisting on a repair that addresses the tummy aches of these non existent beings is rather far from the point. But that is exactly what someone who complains about teaching evolution because of irreducible complexity sounds like. It is exactly what someone who complains about climate models because Malthus disappointed them sounds like.

There is a vast gulf between the science of complex natural systems and the level of understanding of that science among those involved in public discourse and important policy making. This gulf is one of the most significant problems we face in modern society. When an entire political party in the US devotes much of its energy to stand between scientific knowledge and policy making (see: The Republican War on Science), and our population becomes increasingly distant from understanding even basic science (see: Unscientific America: How Scientific Illiteracy Threatens our Future), we are in trouble.

Two very important defenses against the collapse of civilization owing to ignorance (see: Collapse: How Societies Choose to Fail or Succeed) are the science geek and the skeptic. Often but not always they are the same person. These are the people who will vote correctly and encourage others to do so. These are the people who will not let utterly stupid shit pass by in conversations at family gatherings, parties, in classrooms, at bus stops, and on the Internet, without a critical comment. These are the people who may be scientists themselves but more often are not, who push their skepticism out in front of them as the move with alacrity across their social landscapes, mowing down Teh Dumb as they encounter it.

But how does an earnest skeptic "know" a certain area of science well enough to incorporate that science into his or her life? This applies to the science geek who is not a scientist as well as to scientists concerned with areas other than their own. How do you know the greenhouse gas theory is correct, and can be used to estimate the link between CO2 and the Earth's energy balance? How do you know that a particular phylogeny of proto whales based on fossils is probably valid? How do you know that this or that hominid walked upright something like a human?

I think there are two ways to approach this. One is to go as far as you can in grappling with the literature, to become a sort of quasi expert on a topic, so that over time you can explain the details to someone else and maybe convince them that you know what you are talking about! As you do this, you'll become more familiar with a set of methods, a set of data, a bit of intellectual history, and so on. This kind of detailed familiarity in one area will allow you to evaluate evidence on a more cursory level in other areas. If you work with the fossil record and learn all you can about a particular question, you'll learn about the importance of levers in mechanical reconstruction, of taphonomy, of cladistics, and so on, and can apply this understanding to other fossil related problems. If you work with a particular aspect of the immune system you'll be able to much more easily absorb and evaluate, and use in your day to day skeptical activities, newly reported research in some other area related to the immune system.

The second approach relies on the first, and that is to learn who to trust, and trust carefully. There are excellent science communicators who can explain things with just enough detail that you get the interesting parts, but who are required (by the need for brevity, by their editors, or other factors) to gloss important details. You need to keep track of the difference between knowledge you've reconstructed from mostly original sources and that which you gathered from more secondary sources. These different sources will provide different levels of information, and also, qualitatively different information. Wrestling with primary literature means understanding methods in a way you will not have to if using only secondary sources. How exactly does a mass spec give you ratios of stable isotopes? What is proton emission tomography exactly? How does Electron Paramagnetic Resonance (EPR) let you observe the interaction between molecules in muscle tissue, and how does that shape how experiments are designed? Secondary sources skip over all this interesting stuff and give the broader picture, allowing you to cover more ground laterally, but with less certainty, and less ability to crush those who seem to be getting it rong on the Internet (or elsewhere).

There is a problem with exploring primary sources. You are not allowed to. Or at least it is not easy. You can't just walk into a lab and volunteer to work there for a few months to learn all about some very specific research area. In fact, it would be very bad if research labs just let random people show up and volunteer, because that could go very badly. But maybe you do have some way of getting some kind of volunteer experience. A good friend of mine has been collecting inverts for the county for years, as a volunteer, and has gained a great deal of understanding of the methods, scientific questions, and particular local issues of water quality in her area. Check around, you will probably find something.

The published literature may be inaccessible to you because of the way that publishers have manage to get a stranglehold on what is often publicly funded research results. But you can access OpenAccess sources such as the Public Library of Science. And, your local library should be able to give you access to at least somewhat older source material, and the current published papers in major journals such as Science and Nature.

Blind trust is great. For bungee jumping. But calculated skeptical trust (in secondary sources) can be developed by dipping now and then into the primary literature and if at all possible, getting your hands muddy. Or scraped up. Or irradiated. Or whatever (depends on the science.) At the same time, embed yourself in a matrix of trustworthy contacts, and use the knowledge you gain from working with primary sources to evaluate all that information that is out there.

The use of trustworthy sources, and the "trustworthing" of those sources will become increasingly important, and difficult, as more and more fake science sites appear on the Internet. This is a growing problem. More about that at a later time.