This is the second installment in what I hope will turn out to be a long series about the scientific method. In this segment I want to give three examples of how the scientific method, which I described in the first installment, can be applied to situations that are not usually considered "science-y". By doing this I hope to show you how the scientific method can be used without any special training, without any math, but to nonetheless solve real problems.
Example 1
In my inaugural blog post twenty years ago I wrote:
The central tenet of science in which I choose to place my faith is that experiment is the ultimate arbiter of truth. Any idea that is not consistent with experimental evidence must be wrong.
This was an adaptation of Richard Feynman's definition of science, given in the opening paragraphs of the first chapter of his Lectures on Physics. Note that Feynman did not write the Lectures. The Feynman Lectures were not written as a book, they are transcripts of lectures that Feynman gave while teaching an introductory physics course at Caltech in the early 1960s. These lectures were recorded, and it is worth listening to a few of them to get a feel for that the original source material sounds like.
It is worth reading (or listening to) Feynman's introduction in its entirety. It is only nine paragraphs, or nine minutes.
If you read the transcript you will see this:
The principle of science, the definition, almost, is the following: The test of all knowledge is experiment. Experiment is the sole judge of scientific “truth.”
Note that the word "truth" is in quotes. Why? One possibility is that these are "scare quotes", an indication that the word "truth" is being used in an "in an ironic, referential, or otherwise non-standard sense." This matters because it materially changes the meaning of what Feynman is saying here. Without the scare quotes, the passage implies that there exists a transcendent metaphysical Truth with a capital T and that science uncovers this Truth. If that is what Feynman intended, then this would contradict what I said in the first installment, that science converges towards *something*, but that something may or may not be metaphysical Truth.
You might be tempted to argue that there is no way that I -- or anyone else for that matter -- could possibly know what Feynman actually meant, but that is not true. We can. How? By going back to the original source material: there is a recording of Feynman actually speaking those words. If you listen to it, you will find that the transcript is actually not a word-for-word transcription of what Feynman said. Here is what he actually said, word-for-word:
Experiment is the sole judge of truth, with quotation marks...
and he goes on from there say some other things that are not included in the transcript. I'm not going to attempt to transcribe them because there are a lot of clues regarding his intent in his cadence and tone of voice which I cannot render as text. But one thing should be clear: the use of scare quotes in the transcript is justified because Feynman specifically said so.
Does this prove that this is what Feynman meant? No. Nothing in science is ever proven. It's possible that Feynman, because he was speaking off-the-cuff, said something he didn't intend. It's possible that he was under the influence of alien mind-control technology. It's possible that Richard Feynman never actually existed, that he was a myth, and all of the evidence of his existence is actually the product of a vast conspiracy. But if you think that any of these possibilities are likely enough to pursue, well, good luck to you because I predict you're going to be wasting a lot of time.
Discussion
I'm going to break down the previous example in some painstaking detail to show how it is an instance of the process I described before.
1. Identify a Problem. Recall that a Problem is a discrepancy between your background knowledge and something you observe. In this case, the discrepancy was the use of scare quotes in the printed version of the Feynman lectures, and the background knowledge that this is a transcript of something Feynman said rather than something that he wrote.
2. Make a list of simplifying assumptions. In this case there weren't any worth mentioning.
3. Try to come up with a plausible hypothesis. In this case there were two: one was that this was somehow a faithful rendering of what Feynman intended, and the other was that this was an editorial embellishment inserted by whoever produced the transcript.
4. Subject your hypotheses to criticism. I skipped that step because this is just a trivial example and not worth asking other people to spend any time on.
5. Adjust your hypotheses according to the results of step 4. Not applicable here.
6. Do an experiment to try to falsify one or more of your hypotheses. In this case, we had the original audio recording, and so we could go back to the source to hear what Feynman actually said. And it turned out in this case that this new data actually falsified *both* of our initial hypotheses. The transcript is *neither* a verbatim rendering of what Feynman said, *nor* is it an editorial embellishment by the transcriber. Instead, it is a faithful rendering of Feynman's stated intentions, indeed arguably more faithful than a verbatim transcript would have been because (and note that here I am once again engaging in a tiny little example of applying the scientific method in a very abbreviated way) he had to work around a limitation of the medium he was using, namely, speech, which has no way of explicitly rendering punctuation.
7. Use your theory to make more predictions. I skipped that step here too.
Example 2
The second example comes from a real incident from when I was in elementary school. My family emigrated from Germany to Lexington, Kentucky, in the late 60s. My parents were secular Jews. I spoke virtually no English. As you might imagine in a situation like that, I was not exactly the most popular kid in school. I got bullied. A lot. It went on for five years until we moved to Oak Ridge, Tennessee, at which point I was looking forward to making a fresh start. I was no longer obviously a foreigner. I spoke fluent English. I was familiar with the culture (or so I thought). I would not have my reputation as a punching bag following me around. So I was rather dismayed when, within a few months in my new home, I was once again being bullied.
Here was a Problem. I had a theory: I was being bullied in Lexington because I was a foreigner, and the culture wasn't welcoming to foreigners, especially not German Jews, who were just half a notch above blacks in the social pecking order. But in Oak Ridge it was not obvious I was a foreigner. I spoke unaccented English, I was white, I never went to synagogue or did anything else to identify myself as a Jew. So why was I still being picked on?
To make a very, very long story short, I began to consider the possibility that my original hypothesis was fundamentally wrong, and that the reason I was being picked on had nothing to do with what I was but rather with something I was doing, and that I was engaging in the same provocative behavior (whatever that might be) in Oak Ridge as I had in Lexington. In retrospect this was, of course, the right answer, but it took me a very long time to figure it out. It's hard enough to think straight when you are being bullied all the time, and it's even harder when you are in the emotional throes of adolescence and puberty. But I eventually did manage to figure out that the reason I was being bullied was quite simply that I was behaving like a jerk. When I stopped acting like a jerk, the bullying stopped. Not right away, of course. Like I said, it took a very, very long time, and I'm leaving out a lot of painful details. But I eventually did manage to figure it out and become one of the cool kids (or at least one of the cool nerds).
The point of this story is that I solved a real-world social problem using the scientific method without even realizing that I was doing it. This happened in junior high school. I didn't have the foggiest clue about the scientific method, hadn't even encountered it in science classes, and even if I had, the idea that it would be applicable to something besides chemistry experiments would have been laughable. It is only in retrospect that I realized that this is what I had done. And by coming to that realization, I have since been able to do the same thing deliberately in my day-to-day life to great effect. I think anyone can do this, especially with a little coaching, which is one of my motivations for putting the effort into writing all this stuff.
Example 3
My third example comes from philosophy, and I'm putting it in here because it's kind of fun, but also because it actually turns out to be a generally useful guide for spotting certain kinds of invalid arguments. The Problem we are going to address is: how did the universe come into existence? (This qualifies as a Problem because the universe obviously does exist, and so it must have somehow come into existence, but we don't know how.)
The standard scientific answer is that we don't know. Something happened about 13 billion years ago that caused the Big Bang (which is more appropriately called the Everywhere Stretch, but that's another story for another time) but we have no idea what that something is. Religious apologists are quick to seize on this gap in scientific knowledge as an argument for God, but that is not what I want to talk about here. (I promise I'll come back to it in a future installment.) Instead, I want to explore a different hypothesis, one which is obviously ridiculous, and talk about how we can reject this argument in a more principled way than to point to its obvious ridiculousness.
The hypothesis goes by the name of Last Thursday-ism. The hypothesis states that the universe was created last Thursday in the exact state it was then in. Before that, nothing existed. The reason you might think otherwise is that you were created with all your memories intact to give you the illusion that something existed before last Thursday when in fact it did not.
Like I said, obviously -- indeed, intentionally -- ridiculous. But just because something is obviously ridiculous doesn't necessarily mean it's wrong. Quantum mechanics seems obviously ridiculous too when you first encounter it, and it actually turns out to be right. So being obviously ridiculous is not a sound reason for rejecting a hypothesis.
Can you think of a more principled argument for rejecting last-Thursday-ism? Seriously, stop and try before you read on. Remember that last-Thursday-ism is, by design, consistent with all currently observed data.
You might be tempted to say that last-Thursday-ism can be rejected on the grounds that it is unfalsifiable, but all it takes to fix that is a minor tweak: last-Thursday-ism predicts that if you build just the right kind of apparatus it will produce as output the date of the creation of the universe, and so the output of this apparatus will, of course, be last Thursday (assuming you get it built before next Thursday). The cost of this apparatus is $100M (which is a bargain if you compare it to what the particle physicists are asking for nowadays).
Here's a hint: consider an alternative hypothesis which I will call the last-Tuesday hypothesis. The last-Tuesday hypothesis states (as you might guess) that the universe was created last Tuesday. Before that, nothing existed. The reason you think it did is that you were created with all your memories intact to give you the illusion that something existed before last Tuesday when in fact it did not.
You could, of course, substitute any date. Last Monday. November 11, 1955. Whatever. Last-Thursday-ism is not one hypothesis, it is one of a vast family of hypotheses, one for each instance in time in the past. And at most one of that vast family can possibly be right. All the others must be wrong. So unless there is some way to tell a priori which one is right, the odds of any particular one of them, including last-Thursday, being the right one are vanishingly small. And that is why we are justified in rejecting the last-X hypothesis for any particular value of X.
Note that this is true even if the prediction made by the tweaked version of last-Thursday-ism turns out to be true! It might very well be that if we build the apparatus described above that it might very well output "last Thursday". But this will almost certainly not be because last-Thursday-ism is true (because it almost certainly isn't), but for some other reason, like that the apparatus just happens to be a design for a printer that prints out "last Thursday", and this has absolutely nothing to do with when the universe was created.
Invisible Pink Unicorns
That last example may have seemed like a silly detour, but you will be amazed at how often hypotheses that are essentially equivalent to last-Thursday-ism get advanced. I call these "invisible pink unicorn" hypotheses, or IPUs, because the canonical example is that there is an invisible pink unicorn in the room with you right now. The only reason you can't see it is that -- duh! -- it's invisible. This hypothesis can be rejected on the same grounds as last-Thursday-ism. Why pink? Why not green? Or brown? Or mauve? Why a unicorn? Why not an elephant? Or a gryphon? Or a centaur? Unless you have some evidence to make one of these variations more likely than the others, they can call be rejected on the grounds that even if one of them were correct, the odds that we will choose it from among all the alternatives are indistinguishable from zero.
IPUs are everywhere, especially among religious apologists. The cosmological argument, the fine-tuning argument, the ontological argument, etc. etc. etc. -- pretty much any argument of the form, "We cannot imagine how our present state of existence could possibly have arisen by natural processes (that is the Problem) therefore God must exist." But "the universe was created by God" is just one of a vast family of indistinguishable hypothesis: We cannot imagine how our present state of existence could possibly have arisen by natural processes, therefore Brahma must exist. We cannot imagine how our present state of existence could possibly have arisen by natural processes, therefore Mkuru must exist. And, as long as I'm at it: we cannot imagine how our present state of existence could possibly have arisen by natural processes, therefore an invisible pink unicorn with magical powers to create universes must exist.
Note that this is in no way proves that God -- or Brahma or Mkuru or the Invisible Pink Unicorn -- do not exist. It is only meant to show why certain kinds of arguments that are often invoked in favor of their existence are not valid, at least not from a scientific point of view.
This sin is by no means unique to religious apologists. Even professional scientists will advance IPU hypotheses. This happens more often than one would like. String theory is the most notable example. It is an almost textbook example of an IPU. String theory is not a single theory, it is literally a whole family of theories, all of which are indistinguishable based on currently available data. Some string theorists will argue (indeed have argued) that string theory can be tested by building yet another particle accelerator for the low, low price of a few billion dollars, and maybe it can. I don't pretend to understand string theory. But the overt similarity with last-Thursday-ism should make anyone cast a very jaundiced eye on the claims being made despite the fact that the people making them aren't crackpots. Having scientific credentials doesn't necessarily mean that you actually understand or practice the scientific method.
[NOTE] The "read more" link below doesn't lead anywhere. It's there because at some point I accidentally inserted a jump break at the end of this article and now I can't figure out how to get rid of it. AFAICT it's a bug in the Blogger editor. If anyone knows how to get rid of this damn thing please let me know.