Science is the most successful knowledge generating tool the world has ever seen. It has majorly impacted the lives of almost everyone on earth. But most people trust, and even use this tool daily, without fully understanding it.

Science is an invented tool, not a fundamental aspect of nature. Yet even the majority of scientists are unaware of its limitations.

In this article I will explain some of those limitations. I will tell you about 5 important assumptions of science.

In my book Entering Darkness, one of the main characters states that science is based on a number of foundational assumptions that cannot be proven. It was meant to illustrate that science lacks complete authority. Science is an important and useful tool, but it’s only as good as the assumptions it is based upon.

Before I go deeper, let me clarify that being based on assumptions is not a bad thing. In fact, it’s a necessary thing. There is no such thing as knowledge which is not based on assumptions, with the possible exception of your own existence (but even that is arguable and depends on how you define existence!).

To decide that the computer or tablet or phone that you are reading these words on has actual physical form, you first have to assume that you are not a figment of someone’s imagination or a computer program, that your senses can be trusted to tell you the truth, that you are not crazy, etc. You get the idea.

There is probably no reason to question any of these assumptions, but they are nonetheless assumptions.

Science is also based on some assumptions that seem, at first glance, pretty solid. But are they?

Let’s look at these 5 assumptions of science one by one.

1. A physical universe exists that operates independently of our perceptions.

Does the universe really operate independently of our perceptions? Quantum theory has raised a pretty strong objection to this idea in the form of the ‘observer effect’, which states that the very act of observing something influences the thing or phenomenon you are observing.

This is not mere conjecture. It’s been experimentally shown and even has a very useful application in quantum cryptography. Communications can be encoded in such a way that any attempts to eavesdrop will cause the behavior of the information to change in such a way that the people communicating will be alerted.

If observing something changes its behavior, the universe is not operating independently of our perceptions. This creates a paradox, because if it is true then we can’t fully trust science, and if we can’t fully trust science then it may not be true.

How’s that for a mind-bender?

But the thing is that we must assume that the physical universe exists independently of our perceptions in order to investigate it. If not we wouldn’t be able to form any conclusions because none of the evidence we obtained could be trusted.

This is a very difficult problem. On the one hand, we have to make the assumption in order to use the tool we call science. On the other hand, the tool is telling us that we can’t make the assumption. If we trust that the tool is giving us accurate information, then we can’t trust the tool.

To better understand the situation, let’s consider a more concrete example. You want to measure a resistor in an electrical circuit using a multimeter. The aim is to determine the actual resistance value of the resistor. If the resistor is working, it will have some finite value. If the resistor is burnt out, it will have a resistance of infinity. Here’s the problem: you don’t know whether the multimeter works, and if it doesn’t work, it will show a resistance of infinity.

You don’t have any additional multimeters. So before you test the resistor, you make an assumption that it is not burnt out. Because if it is burnt out, you won’t have any way of knowing anything about it, since an infinite reading could just mean the multimeter is not working.

You test the resistor and get an infinite reading. Now, do you a) stick with your assumption and decide the multimeter is faulty, or b) decide to trust the multimeter and let go of your assumption about the resistor?

Just to make the parallel crystal clear, the multimeter represents science and the resistor is the universe. If we decide to stick with our original assumption that the universe operates independently of our perceptions then we have to let go of some of our trust in science. If we trust that science is telling us the truth, then we have to let go of our assumption that the universe operates independently of our perceptions. But if we do that then we can’t trust science anyway!

2. The universe operates according to certain laws which can be discovered.

We’ve discovered a lot of laws and many of them seem to hold. Others, such as Newton’s law of universal gravitation, not so much. It works most of the time, but fails under certain conditions, which means it can no longer be considered a law. But until Einstein came up with relativity, universal gravitation was a law. It was considered to work under all conditions.

The question is, will the same thing happen with other laws of nature? Will we at some point discover that they fail under certain conditions? There is simply no way to know.

But for the moment, there is no compelling reason to question the assumption that the universe operates according to discoverable laws.

3. Events have natural causes which can be determined according to natural laws.

Science assumes there is a natural explanation for this.

Science doesn’t give us the means to investigate supernatural phenomena. In fact, it assumes there aren’t any. Think about it. Scientific theories must be explainable in mechanistic terms. Science is all about the ‘how’. But supernatural occurrences don’t have a how. That’s the whole point. If we can explain how something occurs, it isn’t supernatural.

This is a thorny problem which has led to plenty of controversy.

Some scientists have suggested that science does indeed have something to say about the supernatural. All we have to do, they say, is to set up an experiment to test for the effects of supernatural intervention in the natural world.

For example, we can measure the effectiveness of prayer by having a number of sick patients treated by regular prayers from religious people, while a control group of sick patients are not prayed for. Then simply analyze the results and see whether there is an increase in the rate or frequency of recovery by those prayed for compared to those not prayed for.

On the surface this may seem like a simple and effective idea. So let’s examine it a bit more closely.

Let’s set aside for a moment the fact that it’s not possible to have a control group in such an experiment because it is impossible to know who across the entire world might be praying either directly or indirectly (e.g. general prayers for anyone who has a particular illness) for those in the ‘control group’. Let’s also set aside that we have no idea how a supernatural force or entity (e.g. God) might decide who to heal and when to heal them and whose prayers to listen to or ignore on any given day. And let’s set aside that said supernatural force or entity might wish to interfere with our experiment on purpose to prevent positive results so as to ensure that the need for faith does not disappear.

If we set aside all of those things (realistically we can’t and shouldn’t, but just for fun), we still have some big problems to contend with. For example, if we found no difference, we would still need to replicate the experiment a number of times. And then we would have to repeat it with prayers from other religions and other religious sects. And then we would have to design and run experiments to test every other type of supernatural occurrence imaginable.

And let’s say, by some incredibly massive undertaking involving many thousands of scientists, decades of work and billions of dollars of funding, we managed to do that (I’m being extremely generous. Of course it would not be possible. No one can even imagine all the possible types of supernatural occurrences, let alone test for them all).

If we found nothing we might then have a tentative theory that the supernatural does not exist. But we would still have the observer effect to contend with. The observer effect would tell us that the very act of observing might have precluded any supernatural happenings from taking place during our observation period. It would tell us that by using scientific methods to test for the supernatural, we prevented the supernatural from happening because our scientific methods demand a natural explanation.

But what if…what if we actually found the supernatural? What if we got positive results from our experiments?

Well, that wouldn’t tell us anything either. Because we would never be able to rule out the idea that there really was a natural explanation for what we observed, but we simply lacked the knowledge to interpret it.

And that’s why science can say nothing about the supernatural. It simply cannot test for something that it assumes is not there.

4. The laws of nature are constant throughout space and time.

This is another assumption that science itself has brought into question.

It has also been questioned by some pretty famous physicists, including Richard Feynman and Paul Dirac.

But do we really need to make this assumption in order to do science? Not really, but we need to make it in order to trust science.

Here’s why.

Let’s assume, for a moment, that the laws of nature are not constant throughout space and time. They vary. A million years ago, the gravitational constant was double its current value because Newton’s law of universal gravitation was different. Such a discovery wold send cosmology into a state of chaos for a while. Much of what we know about how the universe went from the big bang to its current state would be wrong.

Any scientific law found to vary over space and time would force a reevaluation of all scientific knowledge built on that law. But it wouldn’t necessarily break science. Science might still work within such a context, but it would be far less effective and trustworthy. After all, is a scientific law really a law if it can change? And how would we discover where and when the laws change(d)?

Without being able to directly observe the past or other parts of the universe, a lack of constancy in the scientific laws would create an immense challenge. It’s the very assumption of constancy that allows us to determine anything about the past or about places we can’t travel to. If the laws that apply here and now don’t apply elsewhere and elsewhen, then we lack a basis to interpret information about other times and places.

How can we use radiometric dating to determine the age of a rock if radioactive half-lives were different when the rock was formed? How can we use telescopes to determine anything about distant stars and galaxies if light has different properties in other parts of the universe?

Of course, the size of the problem posed by a lack of constancy in scientific laws would depend on the size of the change in those laws. Tiny variations might throw off a lot of calculations, while large variations might destroy established theories.

5. Two contradictory statements cannot both be true.

This is called the principal of non-contradiction. It’s a staple of classical logic, upon which both science and mathematics are based. But believe it or not, there is a whole branch of logic in which this is not the case. It’s called paraconsistent logic and it allows for contradictions.

Paraconsistent logic is able to overcome the liar paradox, which simply states, ‘This sentence is false.’

Yes, that’s a paradox. Why? Think about it. If it’s true that the sentence is false, then the sentence is true, not false. But if the sentence is true then it must be false. Which means it’s true. You get the idea.

The sentence cannot be false unless it is true and it cannot be true unless it is false. It’s a logic black hole. There is no escape. If you keep trying to make sense of it you will eventually go insane.

But what if the sentence is both true and false? This is a contradiction, but it breaks the paradox. No more black hole. Paraconsistent logic to the rescue! Sanity restored!

But this would cause all kinds of problems for science. It’s a Pandora’s box. If we allow contradictory statements to coexist in harmony, how can we ever prove anything? Or more accurately, how can we ever falsify anything?

The basic requirement for any hypothesis to be scientific is that it must be falsifiable. This means that there must be a way to prove it incorrect. This is very important because there is no way to prove that a hypothesis is correct. No matter how much evidence you gather to support it, there is always a possibility that you will eventually find evidence that goes against it. So in science, the only way to test a hypothesis is to try to prove it wrong. If you fail, then your hypothesis is considered correct until someone else finds a way to prove it wrong.

A hypothesis which can’t be falsified can’t be tested.

The statement ‘flying pink monkeys exist’ is not a valid scientific hypothesis because it’s impossible to prove it wrong. We don’t have the capability to search the entire universe for flying pink monkeys, so even if we search for eternity there will always remain the possibility that one exists somewhere. We can never prove the statement wrong.

On the other hand, ‘flying pink monkeys do not exist’ is a valid scientific hypothesis. Why? Because if we ever find one, the statement is immediately proven wrong. It is falsifiable. Science must always assume nonexistence of things until they are actually found. That’s how science works.

If we were to stop assuming the principle of noncontradiction, then no hypothesis would be falsifiable.

What does all of this mean?

As you can see, it’s not difficult to raise questions about the core assumptions of science.

I grew up in a fundamentalist Christian environment, where Creationism was taught as fact, dogma was rampant, and free thinking was discouraged. Fortunately I later realized the limitations of such an environment and left. I opened my mind to the wider world and tried to develop a curiosity and openness to a wide variety of ideas and beliefs.

When I became a biologist I was disturbed to discover that I’d entered a whole new world of closed-mindedness. It’s less extreme, but it’s there. Just like most religious fundamentalists are not educated in the philosophy of religion, most scientists are not educated in the philosophy of science. They are not aware of the limitations of science.

Many see science not simply as a tool to learn about the world, but as an all-encompassing, near-infallible key to the ‘knowledge of the universe’. They believe that the process of doing science (the scientific method, peer review and verification by replication) is inherently self-correcting, so that all false knowledge will eventually be removed from the scientific compendium, leaving only truth.

Unfortunately the majority of non-scientists seem split between those who, like many scientists, believe that knowledge gained through science is absolute, and those who view science with suspicion and mistrust, and are quick to discount it whenever it is inconvenient or violates their own beliefs.

Let’s ignore for the moment the problem that science is done by fallible and biased humans (this is a big problem, but one which is ameliorated over time as major theories are retested by hundreds or thousands of scientists). Even if all scientific data was obtained and analyzed in a completely accurate and unbiased manner, any scientific theory could still be called into question if any of these assumptions would turn out to be wrong.

Does that mean we should turn our backs on science and stop trusting any knowledge acquired through scientific methods? Of course not. Look at all the incredible inventions science has brought us. Engineers invent things based on scientific principles and they work.

Science works.

But we need to be aware that science has limitations. It is a solid system of knowledge generation, but we should still be skeptical of anything it tells us.

The major takeaway here isn’t that science could be wrong. It’s that alternatives to science could be right too.

Science assumes that there is nothing beyond the natural world. But what if there is?

Science assumes that two contradictory explanations can’t both be right. But what if they can be?

Science assumes that we can’t alter the universe with our thoughts. But what if we can?

Science assumes that natural laws haven’t changed. But what if they have?

Science assumes that the universe operates according to a set of consistent laws. But what if those laws can be broken?

A violation of any of the assumptions of science might mean that some of the major scientific theories are incorrect. Or it might not. It might just leave a gap for other explanations.

Science has given us answers to many of the great mysteries of the universe. But are they the only possible answers?