Whenever someone tells you that they “have a theory,” you’re probably a mix of curious and skeptical, and with good reason. Most often, you’re going to hear a wild tale that’s a mix of true and dubious facts, possible but suspicious links between them, guesses that range from informed hypotheses to extravagant speculations, and a story about how they’ve introduced one (or more) modifications to the standard story of how things are known work. This is almost always followed with an important caveat at the end: that they have no proof of this, as it’s “only a theory.” In common, informal use, the word theory is tossed around as casually as words like:

• hypothesis,
• guess,
• hunch,
• idea,
• or gut feeling,

among others. In other words, when someone shares their pet theory with you in this regard, you might humor them by considering it, but you’d be well within the bounds of reasons to dismiss their “theory” without putting too much weight behind it.

But some people take this informal definition of theory — as a mere hypothesis, guess, or idea — and apply it to our most vaunted of scientific theories as well. They’ll argue that Darwinian evolution, the Big Bang, gravitation, and quantum mechanics are just “theories” as well, and therefore shouldn’t be taken to be facts. They’ll argue that theories were overturned in the past, and so they’re justified in not believing today’s theories that are well-accepted as part of the scientific consensus. Some will even point to now-discredited scientific ideas, like the theory of phlogiston, Jungian psychology, or the Lamarckian theory of evolution, or to speculative ideas that have “theory” in the title, like Grand Unified Theory or string theory, to argue that many of our best, currently-accepted theories will eventually turn out to be wrong.

But this is a fundamentally dishonest argument, as any scientist will be quick to realize. Here’s what everyone should know about what the word “theory” truly means: in everyday conversation, in scientific circles, in mathematics, and beyond.

The colloquial use of “theory” is just an “idea”

In our everyday life, the word “theory” that we hear bandied about is simply the same as the word “idea.” We make no distinction between whether that idea is:

• a good one that has merit or a bad one that is contradicted by reality,
• an old one that was long ago formulated by someone else or a novel one,
• a valid one that has successful explanatory power or an invalid one that makes incorrect predictions,
• or a scientific one that can be tested and either verified and validated or contradicted and ruled out, or a non-scientific, unfalsifiable one.

In our everyday life, we all have that friend, neighbor, or relative who has rather kooky ideas — something akin to “the Moon is a hologram” or “a secret race of lizard people is controlling the Earth” — and likely refers to their idea as a theory, just as surely as scientists refer to Einstein’s relativity as a theory. The much-ballyhooed idea of panpsychism, although often called a theory, falls into this category of an untestable notion with no evidence behind it.

In all of these cases, we’re only talking about an idea: the initial germ of a hypothesis. Hypotheses may yet grow into fully-fledged scientific theories, and indeed what eventually become scientific theories typically start off as mere hypotheses or ideas, but simply having an idea doesn’t mean you have a good idea, a valid idea, or even a scientific idea, much less a scientific theory. In order to go from an idea to a scientific theory, there are multiple ways the idea in question must be scrutinized.

Touted as UFO lights, this 2011 photograph over Liverpool does, in fact, document a series of lights in an oddly unnatural-appearing configuration, but this is by no means compelling evidence for aliens. In fact, it’s simply an image looking up at the lights of Radio City Tower under dark, foggy conditions, which does not constitute scientific evidence for either UFOs or aliens.

Is the idea plausible? Based on what we already know, if this idea turned out to be correct, would it undermine or conflict with observations, measurement, and data that we already possess? If so, then the idea — at least, the idea as it currently stands — is not plausible, and that’s a non-started for it becoming a scientific theory.

Does the idea have explanatory power? Based on the idea you’re considering, are there phenomena, measurements, or observations that are presently not explained by what we know, that could be explained by this new idea? If not, then this idea has no hope of becoming a scientific theory, as “predictive power” is the key reason and motivation for the formulation of scientific theories in the first place.

Is the idea testable? This is, in many regards, the biggest one: is there some definitive test that can be performed to test this new idea? Is there some way we can compare this idea to the currently favored, prevailing scientific theory it’s seeking to supplant or replace, where we can put these predictions side-by-side, where they’ll differ from one another, and where nature itself will be the arbiter? If not — if this idea is not scientifically testable — then there never was any scientific merit to it at all. (It doesn’t have to be a “practical with current technology” test, but it does have to be a “possible given the constraints and physical limitations of our Universe” test.)

There are other facets to this idea that might appear nice on the surface — whether the idea is elegant, natural, simple, etc. — but it’s really only these three points of plausibility, explanatory power, and testability that make an idea scientific or not. Unless you can test and evaluate this idea based on the evidence you can gather, the measurements you can make, the experiments you can perform and the observations you can undertake, it is not a scientific idea.

This four-panel animation shows the individual galaxies present within Abell 2744, Pandora’s Cluster, alongside the X-ray data from Chandra (red) and the lensing map constructed from gravitational lensing data (blue). The mismatch between the X-rays and the lensing map, as shown across a wide variety of X-ray emitting galaxy clusters, is one of the strongest indicators favoring the presence of dark matter. The Bullet Cluster, as well as other galaxy clusters, exhibit similar features.

All scientific theories exist within a “framework” for exploration

There are a great many ideas that never make it to the “accepted, valid scientific theory” state, but that nevertheless wind up with the word “theory” attached to them. That’s because, well before something is sufficiently tested to know whether it’s valid or invalid, we have to tease out its consequences so that we know what to look for in terms of evaluating it. This comes during a “theory development” stage. Science, after all, is both:

• a body of facts, including the full suite of all relevant data that’s ever been gathered about a particular topic,
• and the process of exploration, where we go down experimental and observational paths whose destination is uncertain, and where nature has ample opportunity to validate and confirm your ideas, to invalidate or refute your ideas, or to surprise you entirely.

That’s why it’s so important that we create frameworks where these various ideas can be tested up against one another, directly. Famous examples where we can do this include:

• the theory that the Earth is flat versus the theory that the Earth is spheroidal,
• the theory that our Solar System is geocentric versus the theory that it’s heliocentric,
• the theory that evolution is based on the use-and-disuse of traits (Lamarckian) or based on the combination of mutations/variations and natural selection (Darwinian),
• or the theory that the Universe exists in a steady, unchanging state versus the theory that the Universe began with a hot Big Bang and has been expanding, cooling, and evolving ever since.

The key, in all of these cases, is specificity: a scientific theory is specific about its predictions, and those predictions can be validated or invalidated by actual data that is measurable and observable.

One of the great puzzles of the 1500s was how planets moved in an apparently retrograde fashion. This could either be explained through Ptolemy’s geocentric model (left), or Copernicus’ heliocentric one (right). However, getting the details right to arbitrary precision was something neither one could do. It would not be until Kepler’s notion of heliocentric, elliptical orbits, and the subsequent mechanism of gravitation proposed by Newton, that heliocentrism would triumph by scientific standards.

In order to work out the consequences of your initial idea or hypothesis, we need to create a framework for testing the idea/hypothesis against the alternatives. This must be based on observation, measurement, and/or experiment; it cannot be a purely intellectual exercise that occurs only in our imaginations. The ability to conduct that translation — from the sandbox of our minds to the hard reality of experiment and observation — is why a testable framework is so important.

Today, the key observations supporting:

• a spheroidal Earth,
• a heliocentric Solar System,
• a Darwinian perspective on evolution,
• and a cosmological hot Big Bang

are overwhelming, and that’s why those theories have stood the test of time and remain accepted today. The alternative proposals have fallen away into the trashcan of science history, and while contrarians often attempt to revive them in modified form, the prevailing theories have continued to beat back such challenges.

However, it’s important to remember that new theories — and in some cases, modified and resurrected variants of old theories — continuously arise in the scientific literature. The theory that there’s a massive “Planet Nine” out there beyond Neptune, distinct from the more typical, puny bodies in the Kuiper belt or Oort cloud, remains an active area of interest, but until the key evidence supporting its existence arrives (it has yet to do so), this will remain a speculative hypothesis, rather than an accepted scientific theory.

Many planetary systems are known where a massive, distant planet orbits far from its parent star. Here in our own Solar System, the hypothesis has been floated that such an object, a “planet nine,” exists far out beyond Neptune. However, it has not been found, and the indirect evidence supporting its presence has been seriously called into doubt by the results of the OSSOS (Outer Solar System Origins Survey) collaboration.

Only validated models that accurately describe reality better than the alternatives become full-fledged, validated scientific theories

This is what happens to the absolute “best of the best” of ideas in science: they become accepted scientific theories. Einstein’s general theory of relativity is one such example, as is the quantum theory of particles and fields in the Standard Model. This happened in biology for theory of Darwinian evolution (inheritance of traits and the evolutionary mechanism of natural selection), in medicine for the germ theory of disease, in chemistry for the theory of atoms, and in many other arenas: where one idea made successful predictions that were borne out and validated by observation and experiment — by a huge suite of robust scientific data — whereas their alternatives failed in one or more ways.

This only occurs when you have an initial idea that was plausible, with explanatory and predictive power, and that has unique features that differentiate it from other, alternative, rival ideas. This furthermore only occurs when the idea is formulated under a framework that allows observable, measurable, quantitative consequences to be derived from that idea, and when those consequences and predictions are then put to the critical test in every way possible, and passes those tests every time and in every way. When we talk about our accepted scientific theories of the day, this is the space in which they all live.

In biology, organisms inherit traits from their parent organism(s), with the only differences coming from mutations and, if there’s more than one parent, a combination of the genetic information present in both parents’ genomes. Darwinian evolution explains how evolution by natural selection works, but genetics, and more recently, DNA, more fully describe the mechanisms behind that evolutionary process.

There are a great many people — scientists chief among them — who are eager to challenge these accepted theories, also known as the “consensus position” in science. There are even a great many scientists who very, very strongly believe that the consensus position will not hold for long, and that a new, superior theory will come along that overturns that accepted theory. Some of them even think that their pet ideas or hypotheses will be the one to do it.

But there’s a 3-step formula that every successful scientific revolution has followed, and whatever scientific revolution you’re attempting to instigate must follow as well. That formula is:

1. Step 1: Your new theory must reproduce all the successes of the leading theory. Whatever the currently accepted theory successfully predicts, describes, and explains, your new theory must successfully predict, describe, and explain as well.
2. Step 2: Your new theory must succeed where the prior theory does not. Every theory has what can be thought of as a “range of validity,” where within that range, its predictions agree with reality, but outside of that range, either its predictions are invalid, untested, or flat-out disagree with what we observe. In addition to succeeding wherever the prior theory was successful, your attempt at a revolutionary theory must succeed here, too.
3. Step 3: Your new theory must make new, testable predictions that differ from those of the original theory. Not only must you explain everything the old theory could plus aspects of reality that the old theory couldn’t, but you must then show where new, yet-unmeasured predictions of your new theory and the old one differ, and then go out and measure the critical phenomena.

Unless and until you clear all three of these steps, any claims of a scientific revolution are premature.

A mural of the Einstein field equations, with an illustration of light bending around the eclipsed Sun: the key observations that first validated general relativity four years after it was first theoretically put forth: back in 1919. The Einstein tensor is shown decomposed, at left, into the Ricci tensor and Ricci scalar, with the cosmological constant term added in after that. If that constant weren’t included, an expanding (or collapsing) Universe would have been an inevitable consequence.

Einstein’s General Relativity only supplanted Newtonian gravity because it reproduced all of Newton’s successes, explained the advance of Mercury’s orbit in a way that Newton’s theory couldn’t, and then went beyond: predicting the deflection of starlight near the Sun during a total solar eclipse, which was only validated years later. Darwinian evolution by natural selection explains a great deal of what we observe, but the mechanism of inheritance — genetics, which itself was a foundational building block that led to the discovery of DNA and RNA: the underlying genetic code — would go far beyond anything Darwin could have anticipated.

Nearly all of the novel ideas you hear about in science news will never even make it to “step 1” of this three-step formula; that’s how robust and well-established our current scientific ideas are across most fields. Most of the ideas that you read about in the popular press cannot even reproduce the full suite of successes of the currently accepted scientific picture; the ideas are that well-established, and concocting a viable alternative is more difficult than most people recognize.

When someone says, “It’s just a theory” about one of the well-established scientific theories that fall into this category (and yes, this category includes the science of human-caused climate change and global warming), remind them of the important difference between an unsubstantiated idea, guess, or hypothesis, and a robust scientific theory. Although nearly all of our theories are almost certainly not the final, ultimate answer, they are the best, most successful way we have to make sense of the Universe we inhabit. Equating those successes with someone’s uneducated guess is an insult to the very enterprise of human civilization.

Yearly surface temperature compared to the 20th-century average from 1880–2021. Blue bars indicate cooler-than-average years; red bars show warmer-than-average years. Not included in this analysis is data from 2023, which has been recently confirmed as the hottest year yet on record. The warming, on average, is by 0.07 C per decade, but has accelerated in recent years, warming at an average of 0.18 C per decade since 1981.

But what about “string theory” and other “theories” like that?

Sometimes, people will bring up examples of theories that scientists study, but that are still only speculative as far as the evidence is concerned, in an attempt to undermine the successes of established scientific theory. What about:

• string theory,
• grand unified theory,
• the theory of extra dimensions,

or some other idea or hypothesis that has the word “theory” in its name, but that doesn’t rise to the same standards of evidence as our established scientific theories?

There are two good reasons for that type of name, however. One of them comes from science: they are theories in the stage of “theory development,” awaiting testing and confirmation that may or may not bear them out. The other is a different definition of theory: one that comes from mathematics. In mathematics, a theory doesn’t need to reflect reality, but rather occurs where you start with a set of axioms — mathematical rules for your system — and those axioms provide you with a framework to derive the properties of any element within that system. Mathematical theories are often the foundational starting point for what will lead to physical theories, and the sandbox of ideas that theoretical physicists most frequently play in is usually based on mathematical extensions to our presently known, accepted physical theories. A physical and mathematical theory can be related, but there is a distinction between how scientists and mathematicians use the word itself.

This important difference explains why, although both string theory and quantum field theory have “theory” in their name, only quantum field theory is established, validated science. String theory, although it is a robust mathematical theory, remains in the realm of “speculative” on scientific grounds.

This particle chart shows the pattern of charges W (weak isospin), W’ (weaker isospin), g3 and g8 (strong force charges), and B (baryon minus lepton number) within the SO(10) grand unified theory. The pattern has been rotated to show its embedding within the larger E6 group: the group that is expected to arise when string theory scenarios break down (or compactify) from 10 (or 11) to four spacetime dimensions.

Unless your theory has that predictive, explanatory power that has proven itself successful time and time again, over a wide variety of physical tests, there’s a very good chance that the “theory” you’re talking about is a theory only in the colloquial sense of the word. Without clearing all of those necessary hurdles, it isn’t a bona fide, accepted, valid scientific theory. String theory is a theory in the mathematical sense, but until its predictions become testable and are validated by those tests, it cannot be taken to be a definitive description of reality. Same for grand unified theories, theories of extra dimensions, the theory of supersymmetry, and many others. They may yet turn out to be true, but science demands that the evidence supports their validity before we can accept any of them.

So yes, your Uncle Randy’s theory about the lizard people, your brother-in-law Fred’s theory about the Loch Ness monster, and Albert Einstein’s general theory of relativity can all be called theories in our everyday language, but only one of them is an accepted theory in the scientific sense. Randy and Fred have theories that are merely speculative and unsubstantiated ideas, where as Albert’s theory represents the pinnacle of modern scientific knowledge. The next time someone tells you that Einstein’s relativity is “just a theory,” you’ll be able to fully appreciate precisely how powerful, and important, being a scientific theory truly is!

This article was first published in August of 2023. It was updated in November of 2026.

This article A “theory” has a special meaning in science is featured on Big Think.

Source link