Chapter 9
What is Science?
Science is a method of investigation. That’s all it is. However, this method of investigation has produced so much accurate, useful knowledge that societies using this knowledge have been able to live in a state of health and well-being that was unheard of in previous centuries.
Scientific investigation has a number of important components, both in how the research is done, and how it is reviewed.
Observation
Science starts with careful observation. Only things that are actually seen, actually heard, or can be detected by specific instruments can be used. The instruments themselves must be available for public scrutiny, so that all aspects of scientific observation are done in a transparent manner. Transparency is of the utmost importance, since results are not considered valid if they can’t be reproduced by others.
Measurement and Quantification
You can’t compare one thing to another without measuring them both. We figured out the orbits of the planets around the sun by doing numerous painstaking measurements of where the planets were in the sky, at what angle, in what places and at what times, for many years on end. But once we had figured out how the orbits were shaped and how much time they took, we were able to figure out exactly where the planets would be in the sky for years and even centuries to come. Astronomers today can tell you exactly where the planet Jupiter will be in relation to the earth, the sun, and to the other planets in our solar system in ten thousand years’ time. Easily. In science, measuring things, and figuring out how to measure them (this is called quantification) is fundamental to what we do. Once some phenomenon has been measured carefully and repeatedly, then scientists can try to figure out what patterns there are, if any. The patterns can be used to make predictions about the results of future observations or future experiments. If these predictions have measurable outcomes, then experiments can be done or observations made, and their results can be assessed in a meaningful way.
Hypothesis and Prediction
A hypothesis is an idea that can lead to a prediction. To do an experiment and have solid experimental evidence, you need to make a specific, testable, quantifiable prediction that is based on your hypothesis.
Experiments and Controls
When you do an experiment, you are testing your prediction. To adequately test a prediction in this way, you must create the circumstances in which you can find out if the prediction came true, or not. This requires two careful procedures. The first is removing all other factors that might influence your results, so that you are testing only the one, single prediction. The second procedure is to do the entire experiment twice. The first time, you include the single factor that you are testing. The second time, you don’t include it, so that you can see what your results would look like in the absence of the tested factor. That way, whatever difference you see between the first run and the second one can be reasonably attributed to the factor you were testing, and not just to the fact that an experiment was being done. This is called doing a controlled experiment, and it’s the basis for a whole lot of science.
It is also sometimes necessary to repeat this whole process many times, just in order to reduce the probability that some of those outcomes were flukes that had nothing to do with the hypothesis that you were testing. A single experiment may take years to do.
This is why science often seems to take so long, and why it is so laborious. Day-to-day science is painstaking and goes at a snail’s pace, but the small bits of reliable knowledge that have been gathered this way have accumulated to the point now where our knowledge about the material world is vastly superior to our knowledge at any other time in human history. Science is very slow, but very effective.
ID is not done this way. ID proponents simply refuse to make testable hypotheses about the material world, and they certainly do not test them. So without predictions, experiments, or quantifiable results, ID cannot realistically claim to be science.
It’s not just me who says that ID doesn’t make predictions. Here’s William Dembski on the subject:
Yes, intelligent design concedes predictability.34
Statistical Analysis and Conclusions
After an experiment has been done, you don’t get to simply say “See! I was right!” You have to analyze your results using statistical tests, in order to make sure that your results didn’t happen by chance. You also have to give the most cautious explanation possible for your results, and not make any claims that your evidence doesn’t solidly support.
Publication in Peer-Reviewed Journals
This is another time-consuming process. After all the experimentation and analysis, you then have to write about your experiment, starting with the hypothesis and the prediction, and including everything about the methods you used and the results you got and the analyses you did. You then need to send this written report to a journal that is run by other scientists. Those other scientists then get to pick apart your work, and figure out every possible way in which you could be wrong. If they don’t like your report, they can refuse to print it. If they think it’s okay, then they may print it in their journal, where it will be read by other scientists, and probably picked apart some more. Sending your report to a journal run by other scientists who know your field is called peer review. People can, of course, publish their results on the internet without any peer review at all, but most people won’t believe results that are published without peer review, and most scientists certainly won’t.
Reproducibility
Reproducibility is a key factor, and perhaps the biggest, most important test in all of science. Once your results have been published in a peer-reviewed journal, other scientists must be able to obtain your results, using the methods that you published in your paper. If repeated, honest attempts by more than one laboratory to reproduce your results fail, then your results will be considered to be invalid.
Many interesting ideas, hypotheses, and theories have been rejected because the results claimed by the original scientists couldn’t be reproduced. For instance, some very exciting initial results that promised a new source of cheap energy through cold fusion were rejected because after repeated attempts, the authors’ results could not be reproduced.
This was not a case of a theory or hypothesis being rejected because it was revolutionary, even though it was. Many people wished the results were true, because they could have solved the world’s energy problems. But the results wound up being considered to be invalid because they could not be reproduced.
By contrast, ID papers do not even have a section of their papers where they describe their methods, so no one else would be able to verify their results. This, all by itself makes ID papers the kind of papers that would be rejected by any peer-reviewed scientific journal.
What Is a Theory?
A theory is a model of how the material world works. It is not just a single hypothesis, or even a bunch of hypotheses. It is a model for how some very large phenomenon works. It should actually generate testable hypotheses.
The germ theory of disease is a great example of a theory. The germ theory of disease simply states that microorganisms are the cause of many diseases. This theory has been extremely helpful to the human race. By understanding that microorganisms are a source of disease, and not some supernatural power, the human race has been able to create safe drinking water, understand safe food-handling practices, and create vaccinations and antibiotics that prevent and cure many diseases. None of these things were possible before the age of modern science.
Interestingly, germ theory doesn’t explain all disease. For instance, cancer is not caused by microorganisms. However, understanding of germ theory combined with not accepting supernatural explanations has allowed us to understand what cancer is, and we are now able to successfully treat many forms of it.
No Supernatural Explanations
One aspect of science that is crucial is that supernatural explanations for the material world are not used. This has proven very useful to science, since it means that scientists keep investigating things, and have been able to explain many things that were previously unexplainable. If we had stopped with an explanation like “God did it,” we would never have cured polio or smallpox, diseases that were once so common and so deadly that people died in the thousands from them; now, many Americans have barely heard of them.
It is not the case that paranormal explanations are rejected out of hand. They are rejected because they don’t provide solid evidence for themselves. Scientists over the years have been willing to investigate anything, including the paranormal. Unfortunately, every time that a scientist does a controlled experiment on people claiming to have paranormal abilities, it turns out that they don’t. Scientists still haven’t found anything in a controlled experimental setting that supports claims of anything supernatural.
Science and the Material World
By definition, science doesn’t know everything. If we did, nobody would do any more scientific research.
But as a means of finding out about the material world, it has an unsurpassed track record.
In the past 400 years since the invention of modern science, the knowledge gained through this method of investigation has improved the health, safety, longevity, and well-being of humankind far beyond anything that our ancestors could have dreamed of.
34. Dembski, Is Intelligent Design Testable? A Response to Eugenie Scott, http://www.discovery.org/a/584.