Science & Nature Archive

Friday, February 19, 2010

Confidence in Scientific Knowledge

Test Tubes & BeakersAs evidenced by one of my recent blog entries, I tend to place a lot of value in science. I think it's the best method we have for answering questions with objectively true answers, and I think we can have a pretty high confidence in the answers it gives us. But, as a few people have recently asked me, where does that confidence come from? Throughout the past, people have had explanations for aspects of the universe that they believed were correct, but have since turned out to be wrong (e.g. the Sun orbiting the Earth). Given humanity's history of failed explanations, shouldn't we expect that many of our current explanations are also wrong, and be a little more cautious in our certainty?

The simplest reason to be confident in science is a pragmatic one - just look at the results. Science as the formalized discipline that we're used to is a fairly recent development. It's only been around a few hundred years, getting started in the Renaissance, but not really coming into its own until after the Enlightenment. But look at how fast our technology has progressed in that short time compared to the previous millenia of human existence. We've invented telescopes, steam engines, automobiles, semiconductors, airplanes, computers, TVs, radio, lasers, vaccines, antibiotics, cures for some cancers. We've sent people to the moon. These accomplishments are all based on knowledge that we've learned through science. It seems very unlikely that we would have been able to accomplish all of that if we didn't have a pretty accurate understanding of reality. Granted, there are other fields of science that haven't yielded practical applications, and possibly never will. For example, understanding the Big Bang may not ever give us any new technologies. However, given the technologies we have developed from other fields, we know that the methods produce reliable results.

Moving away from pragmatism, let's look at how science works. Richard Feynman once said, "Science is a way of trying not to fool yourself. The first principle is that you must not fool yourself, and you are the easiest person to fool." There are all types of ways that we can make mistakes in our reasoning. There's a great article I've linked to before from this site, which does a fantastic job of discussing this: The double-blind gaze: how the double-blind experimental protocol changed science. The article is focused on medicine, but it's applicable to science in general. The article mentions a few of the confounding factors that can affect our reasoning, including the placebo effect, the re-interpretation effect, and observer bias. Wikipedia has a whole list of cognitive biases. A big part of science is recognizing and accounting for all these potential mistakes. Along similar lines, science is not just a search for evidence that confirms your ideas. It's a search for evidence that would disprove your ideas. A big part of science is recognizing when you're wrong.

Science also trains us to think less in terms of absolute certainty, and more in terms of degrees of certainty. If you're being honest with yourself, there's no way to be absolutely certain of anything. It's possible that we're living in The Matrix, or hallucinating, and nothing is as it seems (if this sounds familiar, I've discussed it before). In normal everday conversation however, we tend to ignore those types of outlandish possibilities, and say that we're positive of something, even if technically we mean nearly positive. There are many things we've learned through science that we can say that we're positive are true. The roughly spherical shape of the Earth, the Earth orbiting the Sun, common descent (if not all the exact lineages and mechanisms), are examples of a few of those facts. We should no sooner expect those facts to be overturned than we should expect to wake up on the Nebuchadnezzar fighting alongside Neo. Other things we've learned through science don't have quite as much evidence. Antrhopogenic global warming is an example of this. We can say that we're really darned sure that climate change is happening and that we're responsible, but it's not quite so certain. It would still be really surprising to see AGW turn out to be false, but not earth shattering. You can keep moving down through levels of certainty through things like String Theory, which doesn't really have any evidence confirming it specificaly over other theories, but which is at least consistent with known evidence. If string theory turned out to be false, I wouldn't be all that surprised. You can go even further, and find theories inconsistent with known evidence, such as the supposed link between vaccines and autism, or the aether theory of light. We can be pretty sure that those ideas are false.

In addition to making us think in terms of degree of certainty, science also makes us think in terms of degree of accuracy. Isaac Asimov wrote a good essay titled, The Relativity of Wrong. You should read the whole thing, but here's a great quote from that essay, "When people thought the earth was flat, they were wrong. When people thought the earth was spherical, they were wrong. But if you think that thinking the earth is spherical is just as wrong as thinking the earth is flat, then your view is wronger than both of them put together." An example I've used before is the atom. The current model is the valence shell model, where electrons have a probability of being in particular positions relative to the nucleus. This is an improvement over the Bohr model, where electrons travel in circular orbits around the nucleus and where the orbit radii are defined by quantum mechanics. The Bohr model was an improvement over the Rutherford model (or Solar System model), where the electrons orbited the nucleus, but quantum mechanics wasn't incorporated to predict the orbit radii. The Rutherford model was an improvement over the plum pudding model. And the plum pudding model was at least more accurate than not knowing of the existence of electrons. So, you can see how our explanations have gotten more and more accurate concerning the structure of an atom. Our current model may also be supplanted, but at least we're zeroing in on the truth.

Those are the reasons why we can have confidence in what we learn through science. It's produced results that just wouldn't be possible if the methods didn't work. But it's not simply a matter of thinking that everything science reveals is absolutely right - it's recognizing how science works, what explanations are most likely to be true, and how close we should expect those explanations to be to the actual truth.

Friday, January 29, 2010

'Scientific' Facts

MicroscopeSometimes, a term that you've heard your whole life suddenly seems strange, That's how it is for me and 'scientific facts'. When you think about it, that phrase seems a bit redundant. If something is true, it's a fact. It's that simple. It doesn't matter how you came to know it. If a statement lines up with objective reality, it's a fact.

What does it add to describe a fact as 'scientific'? I guess the first thing is to understand is what's meant by science. Generally, there are two related meanings to the word. The first is that it's a method. We should all know this method from grade school - come up with an explanation, gather evidence to test the explanation, refine your explanation, and repeat. The second is the body of knowledge we've learned through that method. But the thing is, everything that has an objective answer can be examined through science.

Consider an example. Some would consider the Earth orbiting the sun a 'scientific' fact. We as humanity may have learned about it through science, and we as individuals may have learned it in science class, but it doesn't change the fact that it's true. It's not as if the Sun used to orbit the Earth until Galileo came along. Can't we just call it a plain old fact?

There are a couple reasons I bring this up. One is for the people who like to point out that science can't tell us anything with absolute certainty, and therefore science doesn't deal in facts (like this exchange I had). When you consider things like solipsism and Last Thursdayism, you have to grant that for fact to have any meaning, it must mean very high level of certainty, and not 100% absolute certainty. Going by that definition, science certainly does deal in facts.

The other is for the people who think of science as something separate, as not really describing things as part of their world. To them, it may be a 'scientific' fact that evolution occurs, but but in their world, science is wrong, so describing evolution as 'scientific' means it may not have actually occurred.

Oh well, I'm not be expressing myself as clearly as I'd like, but it's late on a Friday, and I'm about ready for some supper and a beer. I guess the main point I'm trying to get across is something I already said in the first paragraph. Calling something a 'scientific' fact is redundant. Statements are either true or not, and if they're true, then they're facts. Since we can study everything with an objective answer through science, it really doesn't add anything to describe any facts as scientific. If they're not scientific, they're not really facts to begin with.

Added 2010-02-01 I thought about this a bit over the weekend, and realized that that last sentence might come off as a bit smug. So, I thought that maybe I should list a couple examples.

As the first example, consider the claim that Hawaii is the 50th state of the U.S. To look at this scientifically, we need to gather evidence to support that claim. We could start off by looking at current legal documents, which show that Hawaii is definitely a state. We could move on to archived documents, and find the Hawaii Admission Act, which shows when Hawaii became a state. We could move on to find documents of when each of the previous 49 states became states. We could study newspaper articles from each of those periods for additional confirmation. After studying all that evidence, then we could say that it is a 'scientific' fact that Hawaii is the 50th state of the U.S.

Next, let's move on to something that some would think was a bit more subjective. Consider the claim that I love my wife and daughter. To test this, people could observe my behavior around my family, and the actions I commit in relation to my wife and daughter. They could study my involuntary facial expressions, to see how I react around them. They could observe my behavior when they're not around, looking for signs of loneliness, or observing how I talk about them. So, even the claim that I love my wife and daughter can be considered to be a 'scientific' fact, since we can use the scientific method to investigate it.

That's what I mean when I say that all facts worth talking about are scientific. Sometimes, we only practice rudimentary forms of the scientific method to determine their veracity, but, at least in principle, the scientific method can be applied to them.

Thursday, November 5, 2009

Ray Comfort - Still Ignorant on Evolution

On the Origin of Species - The Ray Comfort EditionWow. Just, wow. I know I've talked about Ray Comfort more times on this blog than is healthy (for example - here, here, here, here, here, and here), but now, not just is he publishing his drivel on his own, making scam websites, or getting followers to put the equivalent of junk mail into books at the book store. Now, he's been published in a blog on the U.S. News and World Report website, and boy is it ignorant.

The background of this article is this. Ray Comfort is publishing two versions of a reprint of Darwin's Origin of Species, along with an introduction in each version. The first version was abridged, and the introduction was made publicly available on the web. After the negative publicity it received, Comfort made his second version unabridged, and supposedly with a modified introduction. To give an idea of the introduction, here's how Comfort himself described it (be forewarned - there are many falsehoods and examples of bad logic in just these two paragraphs*).

This introduction gives the history of evolution, a timeline of Darwin's life, Hitler's undeniable connections to the theory, Darwin's racism, his disdain for women, and his thoughts on the existence of God. It lists the theory's many hoaxes, exposes the unscientific belief that nothing created everything, points to the incredible structure of DNA, and the absence of any species-to-species transitional forms.

It presents a balanced view of Creationism with information on scientists who believed that God created the universe—scientists such as Albert Einstein, Isaac Newton, Nicholas Copernicus, Francis Bacon, Michael Faraday, Louis Pasteur and Johannes Kepler. It uses many original graphics and "is for use in schools, colleges, and prestigious learning institutions." The introduction also contains the entire contents of the popular booklet, "Why Christianity?"

Towards the end of September, Dan Gilgoff posted an entry in his God & Country blog on U.S. News & World Report describing Comfort's book (the first version). After all the feedback Gilgoff got for that entry, he decided to revisit the issue. He set up an online debate between Ray Comfort and Eugenie Scott, the executive director of the National Center for Science Education. The debate consisted of four posts in total - Comfort's original argument, Scott's original argument, Comfort's response to Scott, and finally, Scott's response to Comfort.

I guess there are several ways I could have addressed this in a blog post, but I've decided to focus on Comfort's second post. That one struck me as so out and out ignorant, that it seemed a ripe target. I encourage you to read Scott's response first, but I thought I could supplement what she already wrote.

Continue reading "Ray Comfort - Still Ignorant on Evolution" »

Wednesday, September 23, 2009

If Evolution Isn't Directed, Why Is Life Now More Complex Than in Ancient History?

A common misconception about evolution is that it has a goal, that organisms evolve from lower to higher forms. This is sometimes referred to as the Ladder or Progress, with primitive forms at the bottom and more advanced forms towards the top of the ladder (and commonly, being the self centered species that we are, with humans on the top rung). This isn't true. Evolution has no direction. Organisms adapt to fit their local environment in whatever way works best. But if evolution has no direction, why is it that life now is more complex* than life from billions of years ago?

This really is pretty easy to understand once you give it a little thought. Let's use distance as an analogy. There's an old saying, that the journey of a thousand miles begins with a single step. You can't walk from Los Angeles to New York instantly. It takes many, many small steps (literally in this example) to get there. If we consider Los Angeles to be simple, and New York to be complex, then at any point on your journey, as you've increased your distance from LA, you've increased your complexity. And it's obvious that you can't get to a certain point of complexity until you've already taken all the previous steps leading up to it. You can't just instantly go from simple to complex.

But a journey still implies direction, and I've said evolution doesn't work that way. Evolution is more like a drunkard's stagger. If you have a drunk that starts out in LA, and let him wander aimlessly with no particular destination in mind, he may eventually end up in New York, but it definitely wouldn't be a straight line. He may just as likely never make it to New York, and never even leave LA. To extend the analogy further, he may end up in Seattle. He'd still be a long way from LA, but in a completely different direction. Squids, for example, are remarkably complex, but they took a different path to their complexity than us vertebrates, and their resulting complexity is different from ours.

Below is a graph that roughly illustrates this in an evolutionary context. It starts off at zero, and for every step, it goes up or down by a random amount** between -0.5 and 0.5. After every 10 steps, it splits, and each new 'lineage' then varies in that same manner. This was carried on for 40 steps, resulting in 8 lineages by the end.

Random Distribution Simulating Evolution of Complexity

Remember, this is all random variation from a starting point at zero, going in small steps. After 40 steps, one lineage had varied to more than 3.4 away from zero, while other lineages didn't vary very much away from zero at all. If this was representing complexity, and if the steps were assumed to be thousands or millions of generations, it demonstrates how complexity can evolve slowly from simple beginnings, without any conscious aim toward increased complexity. (As I said, this is only a rough illustration of evolution. Evolution is driven by more than just random variation, and the divergence of lineages isn't as predictable as that.)

Evolution really does sometimes decrease complexity. For an intimate example, consider what you're sitting on - a nice smooth posterior. Some time millions of years ago, our ancestors lost their tails, a complex feature with muscles, bones, and tendons. Their lifestyles were probably such that a tail just didn't really do that much good, so there was no reason for natural selection to maintain it. And now, we have backsides that are less complex than our simian ancestors.

*'Complex' is actually a little hard to define. How exactly do we mean complex? Number of genes? Number of specialized cell types? How do we even differentiate specialized cell types? (more info) This seems like one of those problems where we know it when we see it. I think most people would agree that a mammal is more complex than an earthworm, even without a quantitative definition of that complexity.

**Technically, a pseudo random number generator was used (Visual Basic). For this application, that's close enough to truly random.

Wednesday, June 3, 2009


I had an interesting experience at work last Friday. Some time in the early afternoon, we started noticing some bees around the entrance to a trailer. We hadn't seen any yet this summer, so I poked around to see where they were coming from. I saw a few going in and out of a hole in the liner under the floor, so I ran home to get some wasp poison, and sprayed around the hole. As the afternoon wore on, more and more bees kept showing up. I kept spraying for a while, but eventually gave up. By the time I left to go home, here's what it looked like (I gave two different resolution options to view these bigger):

Bees View 1
800x600 (92.4 kB)
2816x2112 (504 kB)

Bees View 1
800x600 (104 kB)
2816x2112 (514 kB)

Bees View 1
800x600 (104 kB)
2816x2112 (491 kB)

And here are a couple videos of it that I uploaded to YouTube. The first one's a bit long, but you don't have to watch the whole thing. At the end of the second one, I got scared when I felt something land on my arm, but I didn't get stung.

By Sunday, the bees were no longer on the step, but had taken up residence under the trailer. We had to call in a bee keeper to get rid of them. He came in after dark on Sunday (so that the bees would be in the hive) and worked until midnight. He said that there must have been at least 20,000 bees. He didn't think they'd been there very long, that they'd probably just arrived.

That is, by far, the most bees I've ever seen in my life.


Selling Out