#25-6 In Defense of Science, Part 1

In which the E@L defines Science, the Null Hypothesis, and the Black Swan

Science in the United States is under attack. Funding for the NIH has been slashed. Grants for research on cancer, disease, HIV, coronavirus, climate change and other critical issues have been canceled. University overhead budgets have been reduced to untenable levels (OK, having fought my own University overhead battles, I get this, but I still needed a functional laboratory, with water and lights). These events are so pervasive and abundant, I don’t even need to include links to them.

All of this destruction can be blamed directly on the administration of President Trump and his supporters. Their attitude toward science seems to be that, if they don’t understand it, it shouldn’t exist. But more likely, they are just uneducated, uninterested, and insecure people who hate the fact that other people are smarter than them. And they are killing the golden goose.

The current position of the United States as a (the?) world leader in technology, innovation, medicine, and research is largely the result of the investments that have been made in science by our government. In the United States, over 103 million people caught Covid-19, and 1.2 million died from it. But the number of deaths would have been much greater if we had not invested heavily in a new and untested technology, the mRNA-based vaccine. But few people even knew what mRNA was, and due to ignorance and fear, many of them insisted that it included microchips, or caused changes to our DNA. And the reason for this is that we, as a society, no longer understand science. We don’t teach it to our students, and we don’t educate the public about the importance of scientific research.

In this series of essays, I am going to attempt a very general explanation of what science is, how it works, and why it is important. I will define hypotheses, good and bad, and explain how they are tested. I will describe the process of experimental research, the importance of controls, and the use of placebos. I will explain the difference between science, pseudoscience, and junk science. As an example of the latter, I will explain how invalid scientific methods were used in support of Hydroxychloroquine as a potential coronavirus cure. I will also explain how the efficacy of the mRNA vaccine was tested, accurately and correctly. And finally, I will discuss how humans are really bad at risk assessment, and place way too much emphasis on unlikely events rather than on more likely outcomes.

This series will require three or four posts over the next few weeks, so I expect only my most dedicated readers will read them all thoroughly. But I think you will find it worth your time, and I hope you will persist.

Science Evolved from Ignorance

If the geological time scale were based on a 12-hour clock, humans have been around for only a fraction of a second. If we are going to persist beyond that, we have a lot of problems to solve. And the only way to do that is with science. Not politics. Not philosophy. Science.

But in order to use science, we need a shared understanding of the language of science. Unfortunately, we are not required to learn it in our educational institutions. You can, if you want, but it’s not a requirement, and most of what is taught is inaccurate. How did this state come about?

Early humans didn’t understand much about the way the world worked. They didn’t know what made crops grow, or why it rained. Rather than admit that it was beyond their abilities, they assigned such unknowables to the realm of the gods. At first many, later one or several. They couldn’t live with uncertainty; so made up rules and explanations for anything that wasn’t certain. These beliefs became their religion. We have whole books about them. Greek and Roman Myths. The Bible. The Koran. The Bhagavad Gita. None of these books were actually based on any empirical observation of the world. They attempted to explain a world that could not be understood by humans at the time, because they did not have a basic concept of science. And despite their many prophesies, they did not predict anything that occurred later in history.

But a few people did have concepts of science. Ptolemy, Aristotle, Plato (to a degree). These few began recording and reporting the process of scientific explanation. At the time, it was considered philosophy, and then a subset of that called natural philosophy. But since most people couldn’t read, it didn’t become common knowledge for hundreds of years. Since then, we have had a great expansion of scientific thought.

Thanks to NOAA for this image.

What is Science?

So, what, exactly, is science? Many great thinkers have offered a variety of definitions.

• The systematic classification of experience. George Henry Lewes (1817-1878)

  • Ok, so we observe, and we write things down, like an encyclopedia. But it doesn’t imply any analysis or understanding of what is happening.

• Nothing but trained and organized common sense. Thomas Henry Huxley (1825-1895)

  • This explanation is a bit too simple. It implies that most people have common sense (arguable), and that a few can learn science if they undergo rigorous training. But I don’t think it’s comprehensive.

• What you know. Philosophy is what you don't know. Bertrand Russell (1872-1970)

  • Spoken by a philosopher, this is a bit too glib. We don’t have to “know” something to call it science. And we can’t know everything, which eliminates a big chunk of it. And what we do “know” is always changing.

• A series of judgments, revised without ceasing. Pierre Emile Duclaux (1840-1904)

  • This, I think, gets at the heart of it. Science reveals information, which we evaluate, and judge. If it is acceptable, we elevate it to knowledge. But we must continuously reexamine that knowledge to see if it needs to be altered, adjusted, or improved.

Then what is science? According to the late Carl Sagan, Science is

• a Process of inquiry

• that uses testable, refutable hypotheses,

• and accepts the results without mystical intervention.

Still, many people do not understand this. And one of their primary misconceptions is the belief that science is a process of trying to prove something. As in “Prove that this (whatever) is safe before we eat/take/ingest/inject/wear/breath/drink it”. But that is not how science works.

Can science prove anything? NO. Science can only disprove. For science to operate, we have to start with observations. Then, we posit a potential explanation for what we observed. This is called a Hypothesis. Then we must test the hypothesis to determine if it is true. But not all hypotheses are created equal. We cannot prove a hypothesis, but in order to disprove it, it has to be expressed in a way that we can refute it. That is, it must be falsifiable. This is what I call Science Rule No. 1.

Hypotheses: The Bad, the Null, and the Ridiculous

Here is an example of a (Bad) hypothesis: There’s a monster under my bed. Is this testable? How? Of course, you are thinking “Look under the bed!”. But what if we do that, and we don’t see any monsters? Does that mean they do not exist? Can we prove this hypothesis? Is it falsifiable?

Which leads us to Science Rule No. 2: Absence of evidence is not evidence of absence.

Our hypothesis is not a good one, because it is not falsifiable. For that, we need to restate it as a Null Hypothesis. As in, There are NO monsters under my bed. Is this testable? Is it falsifiable? How? Look under the bed again?

What if we still don’t see any monsters? That result is called a “False Negative”. Can we then assume the Hypothesis is true? Yes, tentatively. There are no monsters, for the time being. But what if we do see one? Can we reject the Hypothesis? And if so, then what?

Proponents of religious dogma have always tried to defend their beliefs from attack by outsiders. To do this, they stated their beliefs such that it was impossible to refute them. Carl Sagan offers an interesting analogy:

• There’s a fire-breathing dragon in my garage.

  • But I can’t see it.

• That’s because it’s invisible.

  • Suppose we powder the floor and look for footprints.

• But it floats in the air.

  • Suppose we use an infra-red sensor to detect the fire.

• But it doesn’t produce any heat.

  • Suppose we spraypaint it.

• But it is non-corporeal and made of spirit only.

Now what’s the difference between a non-corporeal, floating, invisible, fire-breathing dragon, and none at all? If there is no way to disprove its existence, does that suffice to prove that it exists?

The Black Swan

Suppose we make an observation while watching birds on a lake: A swan is white. From this observation, we make a generalization, as a hypothesis: All swans are white. (This, incidentally, is exactly what Europeans used to think). Is this testable and falsifiable? How? Can we observe all swans in the world? If not, then we cannot falsify it.

Mute Swan, Cygnus olor (Source)

So, let’s rephrase the Question as a Null Hypothesis: There are no non-white swans. Is it falsifiable? YES. We don’t need to examine all swans. All we need is to find a single black swan. (Which actually happened once Europeans discovered Australia and its non-white swans). That would prove that our hypothesis is false, i.e. “NOT all swans are white”, and we can then consider the alternative, which is that “Some swans are NOT white”.

Black Swan, Cygnus atratus. (Image: Jeff Melvaine, © Creative Commons). Literally from the Australian Museum

What if we see a dirty swan and think it’s black? This is a mistake known as a “false positive”, or Type I error.

Finding the black swan was unexpected and expands our concept of swans. But, according to author Nassim Nicholas Taleb, it is more important than that. A “Black Swan” event is any rare event that we do not expect, but we should. Such events have three important characteristics:

1. They are rare, so unexpected.

2. They have an extreme impact that can change the course of history.

3. They cause us to create explanations after the fact about why we should have expected it.

In fact, most of the important events in world history, scientific development, and our personal lives were not predictable, and were, in fact, Black Swan events. We spend most of our lives thinking about and planning for normal, recurring, and predictable events, that we do not plan for Black Swan events, even though we know they will happen, and will have major impacts on our lives.

The problem of science is that, for any observation, there are multiple hypotheses that could explain it. How do we choose which one to test? Which one is most (or least) likely? Remember that all hypotheses must be refutable, i.e. falsifiable.

In order to evaluate multiple hypotheses, we must have a way to compare them. That requires the use of probability, which brings our science up to the last two centuries, a time when probability theory was developed and introduced into the evaluation of science.

But that is a more complex subject which will be explored further in the next post.

---

This issue of Ecologist at Large is available to all readers. However, if you would like to support my work with a one-off contribution, click “Buy me a coffee” below.

---

---

Sources

Sagan, Carl (1997). The Demon-Haunted World: Science As a Candle in the Dark (Reprint ed.). Ballantine Books. ISBN 978-0345409461.

Taleb, Nassim Nicholas. 2010. The Black Swan: Second Edition: The Impact of the Highly Improbable: With a new section: "On Robustness and Fragility". Random House Trade Paperbacks. ISBN 978-0812973815.