I’ve written several articles about what science is and how we all function as scientists in our day to day lives. If you haven’t visited my page on science vs. religion, it wouldn’t hurt to go there before reading this entry. (Hint: Look at the page tabs right above my lovely landscape at the top of the page!)
Today I want to address a couple of concepts inherent in the scientific method, along with some of the myths accompanying them — many of which are believed by atheists and theists alike. The two concepts are certainty and limits.
Nearly everybody knows that science doesn’t claim certainty about anything empirical. That’s not really much of a problem in the real world, but it’s often seen as some kind of “fatal flaw” by theists who believe that certainty is something that we ought to have. But the truth — the beautiful truth — is that uncertainty is one of the greatest benefits to both the scientific method and what I’ll call the “scientific mind.” (A person with a scientific mind is someone who tries to apply the principles of science whenever and wherever possible in day to day life.)
The Problem with Certainty — put simply, there’s a fundamental problem with the very concept of certainty. If we’re certain about something — if we are 100% sure without any possibility of doubt — how do we know that we are certain? If you think about it for a second, you realize the impossibility of it all. Unless we can predict the future with 100% accuracy, we cannot be certain that we’re certain about anything. None of us can know that the next second of our existence will present us with the same evidence we see right now.
Science readily admits this, and doesn’t fret over it. Instead, it plays the hand it’s been dealt. “Ok. I can’t know anything for certain. But how close to certainty can I get?” Some things — like the cause of the Big Bang — are admittedly speculation, and though some scientists are dogged supporters of their own theories, there is no pretense of certainty. Other things, like evolution and gravity, are so close to certain that there’s no reason to doubt them. Scientists proceed as if they were 100% certain, assuming evolution and gravity in all experiments which involve either of them. But even this is not real certainty. Any good scientist will admit that despite the overwhelming evidence for both gravity and evolution, if tomorrow brings stronger evidence for something else, they’ll change their minds.
Science makes certain assumptions about the universe. But these assumptions are not certain either. Scientists assume the future will be like the past for the purpose of prediction. But if it turns out that the future is not like the past, scientists will examine the data and try to figure out why. They’ll say, “Wow… that was really a surprise. Let’s see if we can figure out what happened to change the laws of the universe!”
The beauty of the scientific mind is that it accepts and even embraces uncertainty, so that any event could be the catalyst for a change of opinion. Science is the methodological analog to our greatest survival skill — our adaptability.
It is not the strongest of the species that survives, nor the most intelligent, but rather the one most adaptable to change. — paraphrase of Charles Darwin by Leon C. Megginson
Science is the application of intelligence to adaptability. How much more beautiful an invention could we hope for? For hundreds of thousands of years, humans have adapted enough to survive. But our survival was meager, and often depended as much on luck as anything else. It is estimated that at one point in our evolutionary history, our population was so low that today, we would qualify as an endangered species. But no longer. Once we learned to apply our intelligence to the task of adapting to our environment, our population exploded. Our lifespans more than doubled. Our infant mortality rates plummeted to near insignificance. (Of course, we face a new adaptive problem. How do we adapt to what is now a severe overpopulation problem?)
Certainty and the Daily Scientific Mind
Some people who are unfamiliar with the philosophy of science will balk at the idea of embracing uncertainty. If nothing is certain, how do we ever decide on anything? How do we believe anything? Doesn’t uncertainty inevitably lead to nihilism or solipsism? In practical terms, this kind of thinking is unnecessarily fatalistic and wrong-headed. We don’t need certainty to function effectively and happily. All we need is a combination of probability and intuition. How close to certain are we that X is true? How certain do we need to be before we proceed as if X is true?
We all do this unconsciously every day. The same kind of reasoning works for the mundane and the life-altering decisions we make. Are we more likely to get a better combination of price and quality at Publix or Kroger? Unless we’re stretching out our last ten dollars, it’s not critically important, so even scant evidence towards one store or the other will probably be enough for us to choose. Once we choose, our opinion can be easily swayed in the other direction by equally scant evidence. We really don’t know, and we just “go with it.” No biggie.
But what about my spouse? Is she cheating on me? Unless I’ve got pretty damn strong evidence, I’m probably not going to file for divorce. That’s a really big life change, and it would be foolish of me to hire a lawyer because of one call at 11 PM from “Unknown Caller.” Intuitively, I know that my degree of certainty must be much higher for more important decisions, and generally, intuition is accurate enough that I don’t do too many completely boneheaded things.
And the beauty of it all is that if I am a science minded person, I’m always open to the idea of changing the degree of certainty requirement. Anything is subject to change, but change is a product of evidence. If there is a simple mantra for living a science minded life, it goes something like this: Believe nothing without evidence. Act with strength of conviction equal to the strength of the evidence. Search for and crave new evidence. Learning one is wrong and changing one’s mind is the highest achievement of a scientific mind.
Scientific studies have limitations. This isn’t a flaw. It’s just a reality. Armchair philosophers and neophyte science students often pride themselves on smugly pointing out what a particular study doesn’t demonstrate. Of course, even in the ranks of the scientific elite, it’s possible to overstate the significance or reach of an experiment, but generally speaking, all peer reviewed experiments have a section detailing the limits of the work.
Limits become trickier as the subject of an experiment becomes more complicated. Most individual studies are very restricted in what they address. This is actually a good thing because attempting to study too many things at once usually just leads to messy data. Isolating one or two variables is the best way to discover precisely how they work. If a particular phenomenon has a thousand variables, it might take hundreds of scientists and thousands of experiments to have enough individual studies so that someone can compile a meta-study, which attempts to synthesize all the individual experiments into a cohesive description of the entire phenomenon. Simply put, most complex phenomena are understood through amassing corroborative evidence from multiple experiments.
How do scientists proceed when the data is incomplete and they can’t say for certain how a complex system works? Generally speaking, they make educated guesses and then test them. That is, once a few experiments are finished, scientists may think they see a pattern, and may make a guess as to how the entire system works. They design an experiment: If this is the way things work, then I can do experiment X and get result Y. They proceed as if they understand the whole system, and expect certain results which will reinforce their guesses.
But good scientists do not let their expectations muddy the waters. If the results are not what they expected, they first look for flaws in the experiment which would have led to faulty data. If the experiment is valid, they start trying to form a new theory to account for the unexpected results.
If there’s one thing that causes more trouble for scientists than anything else, it’s competing theories. Unfortunately, not all questions have answers right now. Even worse, some questions demand courses of action right now even though they don’t have definitive scientific answers. Global Warming is a good example. It’s beyond scientific dispute that humans are having an effect on the environment, and that we are spewing billions of tons of greenhouse gases into the atmosphere. There is also no scientific dispute that the levels of these gases are higher today than they were fifty years ago. Unfortunately, our understanding of the atmosphere and how it interacts with geologic, cosmic, and biological factors is far from complete. We simply can’t predict the weather fifty years into the future.
The implications of global warming are significant enough that IF some of the predictions are correct, THEN we need to be taking action right now. But we simply can’t be sure. We don’t have enough data to make a scientific case for exactly what will happen if we continue to emit greenhouse gases at our current rates of increase. To be sure, there are scientists who will swear that their theories are conclusive enough to demand action. (I agree with them, but then, I’m not an ecologist, so that doesn’t mean much.) But there are others who say that there is enough compelling evidence for other theories to warrant further research before any action is taken.
It is in cases like this that we simply have to take our best “educated guess” and go with whichever side makes more sense from a risk/reward point of view. In the best case, we do so with a certain amount of trepidation, and with promises to evaluate our action based on initial results. In the case of global warming, we could test our theories by examining a decade of significantly reduced emissions. Do the changes we see match predictions from the doom-sayers? If so, then we should continue on this course. If not, it’s time to re-examine the evidence and refine our tactics based on the new evidence.
Everyday Competing Theories
How does this apply to the everyday life of a scientific-minded person? Simply put, when there are competing theories, we go with the one that makes the most sense, and if there’s any way to hedge our bet, we do so. In other words, we try not to throw all our eggs in one basket when there are a lot of attractive baskets.
In day to day living, there are a couple of thinking errors that pop up frequently. First, day to day living often involves scientific observations of human psychology and sociology, which are two of the most complicated systems we study. There are LOTS of variables, and the legitimate study of the human animal is still in its infancy. There’s a lot we don’t know. So there are a lot of competing theories out there.
Second, in day to day living, we are almost always heavily personally invested in our own theories and beliefs. Our reputations, our relationships, our sex lives, our incomes… all of these things hang in the balance when we commit to courses of action, and we do not live in a sterile scientific lab where everyone around us is ambivalently taking in data and searching for truth regardless of its content. We are social animals in a highly competitive and bloodthirsty environment, and being wrong is often very detrimental to us. Cognitive bias is an extremely seductive error to make.
Cognitive bias takes many forms. In fact, I think it’s fair to say that almost all errors in social critical thinking are in some way due to cognitive bias. That is, we are heavily invested in being right, so we observe the data through the filter of being right. We tend to minimize that which disagrees and maximize that which agrees.
A very good example of this is pornography. A few days ago, I linked to a literature review which is a preliminary attempt to determine the validity of claims that pornography is dangerous to society. There are a lot of people who are heavily devoted to blanket generalizations on both sides. The religious and prudish see porn as a cumulative societal ill. That is, any positive aspects of it are far outweighed by the damage it causes — exploitation of women, increase of sex crime, destruction of relationships, objectification of women, etc. Liberals are equally committed to viewing it as a cumulative good. Any negative effects it has are far outweighed by the benefits — healthy sexual expression, liberation of women from patriarchal ideals, exploration of fantasy, keeping relationships interesting and exciting, etc.
Whenever a new study on porn is released, the armies of good and evil trot out their standard responses. The prudes say the study has highlighted a potential benefit while ignoring a greater ill. The liberals say it validates their condemnation of the prudes, but they minimize or fail to mention the limits of the study, either out of cognitive bias or fear of giving the prudes any ammunition.
But in the real world, we must address porn. We must allow it or disallow it. In other words, we must make our best guess based on the available evidence and risk/reward analysis. With porn, this is especially difficult because the competing theories are almost diametrically opposed. On the one side, it is claimed that porn is leading to the literal destruction of our society. The other side swears that freely expressing sexuality is critical in minimizing sex crime and maximizing personal fulfillment. Both sides are offering extreme benefits and costs, and choosing the wrong answer apparently has dire consequences.
But even here, there is an error in critical thinking. These are not the only two competing theories. Porn might not have an appreciable effect on sex crime at all. Men might objectify women equally regardless of the accessibility of porn. The exploitation/liberation of women might be illusory, and have little or nothing to do with porn. Both sides could be wrong.
The science-minded person can only do one thing — explore the available evidence as well as possible and determine if the preponderance of the evidence points to an immediate and dire consequence if porn is allowed. If there is no such compelling evidence, he must continue to search for evidence. This is part of the mantra. Believe nothing without evidence. If he is firmly attached to his personal theory of porn, then he owes it to himself and the scientific community to continue to isolate variables, perform experiments, and contribute to the corroborating evidence which will facilitate a general theory. More importantly, he owes it to everyone to respect the limits of his knowledge. His suspicion that he knows of a terrible ill or grand benefit to porn should drive him to research, not legislation.
Risk vs Reward
I’ve given two prominent examples of competing theories that affect our everyday lives. I want to examine the risk vs. reward thinking on each one to show how (at least from what data I can find) one example calls for immediate action and the other calls for inaction.
- Global Warming: If it is true that global warming is being vastly accelerated by humans, and that it will lead to the destruction of our environment, then we have three choices: Change our behavior and stave off disaster; Change our behavior and it really won’t matter either way; Not change our behavior and potentially cause disaster. In examining the risk vs. reward, the only course that leads to disaster is not changing our behavior. While it may not inevitably lead to disaster, it has the possibility of doing so, and there’s no compelling evidence that reducing our greenhouse emissions will lead to disaster. The reasonable choice is to change our behavior.
- Porn: While the claims that porn are dangerous are being shouted very loudly, there is little compelling evidence that it leads to the drastic ends claimed by the prudes. Depictions of human sexuality are ubiquitous and as old as art. There are no known examples of a society destroying itself with porn. There is some evidence to the contrary. While it’s not compelling enough to “disprove” the nay-sayers, it’s compelling enough to suggest that restricting access to sexual expression may be dangerous for society. So while there’s not enough evidence to de-censor the TV and start streaming BDSM into school classrooms, there is certainly enough to suggest that we won’t be doing any great harm by studying porn more without taking any draconian measures against it.
Hopefully, these two examples illustrate how science generally proceeds with caution and the preponderance of evidence. When a decision is forced, the strength of the decision is mitigated by the weight of the evidence. Once a decision is reached, the first course of action is to examine any new data carefully. In examining competing theories, risk vs. reward reasoning should be the final judge of appropriate (and appropriately tentative) action.
I wonder how much better America could be if our politicians proceeded this way. Instead of arguing over the emotions of teabaggers and aging liberal hippie douches, what if they examined the scientific data for issues like abstinence only education, socialized healthcare, and economic legislation? What if they examined things not from the risk/reward paradigm of their voting constituency, but the empirical truth?
More to the point, what if we as a nation of individuals decided to stay our emotions and commit ourselves to a real search for truth? If we gave up our notions of certainty and learned to embrace the acquisition of new data instead of fearing it, how much could we accomplish in our own lives? This is the promise of rationalism and science. This is why uncertainty is so much better than certainty. This is why it’s so important for the freethinker movement to gain momentum and try to change the face of America.