Science is the pursuit and application of knowledge about the natural and social world through a systematic, evidence-based method. That definition, from the Science Council, captures what makes science distinct from other ways of knowing: it relies on observation, testing, and verification rather than tradition, authority, or intuition. The word itself comes from the Latin “scientia,” meaning simply “knowledge,” but what we call science today is far more specific than that broad origin suggests.
From “Knowledge” to a Method of Knowing
In medieval Europe, “science” covered a sprawling range of intellectual pursuits: grammar, logic, rhetoric, arithmetic, geometry, music, and astronomy all fell under the umbrella. People who studied the natural world were called “natural philosophers,” not scientists. The word “scientist” didn’t exist until the philosopher William Whewell coined it in the 1830s.
Around that same period, the meaning of “science” narrowed. It stopped referring to knowledge in general and became attached to a particular way of gaining knowledge, one that relies on observing the world and testing ideas to figure out which claims hold up. That shift is what separates the modern concept from its Latin root. Knowing something through tradition or reasoning alone doesn’t count as science. The knowledge has to be grounded in evidence you can observe, measure, and check.
What Makes Something Scientific
Several core features distinguish science from other forms of inquiry. The most fundamental is empirical evidence, a term that simply means information gathered through the senses: things you can see, hear, measure, or detect with instruments. As the National Research Council puts it, science is grounded in observations that can actually be made about the world. Questions that can’t be investigated through observation or experiment fall outside science’s reach.
Beyond evidence, science requires:
- Testable hypotheses. A scientific idea must make predictions that can be checked against reality. If no possible observation could prove the idea wrong, it isn’t functioning as science.
- Systematic methodology. Researchers follow structured processes for collecting data, controlling variables, and analyzing results, rather than relying on casual impressions.
- Repetition and verification. One of the ways the scientific community confirms a discovery is by repeating the research that produced it. A finding that only appears once, in one lab, under one set of conditions, doesn’t carry much weight.
- Peer review. Before results are broadly accepted, other experts in the field scrutinize the work for flaws in logic, method, or interpretation. A hypothesis or claim generally isn’t accepted by the scientific community unless it has been published in a peer-reviewed journal.
These features work together as a quality-control system. No single experiment settles a question. Instead, evidence accumulates over repeated cycles of testing, criticism, and refinement.
The Scientific Method as a Cycle
You’ve probably seen the scientific method presented as a neat list of steps: ask a question, form a hypothesis, design an experiment, analyze data, draw conclusions. That sequence is real, but it’s more accurately understood as a loop than a checklist. Conclusions from one round of research reshape the original question or hypothesis, which then gets tested again. Each pass through the cycle ideally gets closer to an accurate picture of how something works.
This iterative quality is important. Science doesn’t claim to deliver final, unchangeable truth in a single study. It treats results as evidence, and with more evidence, understanding improves. That’s why scientific knowledge can shift over time without the process being broken. Updating conclusions based on better data is the system working exactly as intended.
The Goal: Explanation and Understanding
At its core, science aims to produce increasingly accurate explanations of how the natural world works, what it’s made of, and how it got to its current state. That understanding takes two practical forms. Sometimes it’s pursued for its own sake, like investigating how a molecule behaves just to learn about its structure. Other times it’s applied directly to problems, like figuring out how to combine materials into a working light bulb or predicting how likely a child is to inherit a genetic condition.
Both of these paths, pure research and applied research, count as science. The difference is intent, not method. A physicist studying subatomic particles and an engineer testing bridge materials are both using systematic observation and evidence to build reliable knowledge.
The Three Major Branches
Science is typically divided into three broad categories based on what’s being studied:
- Natural sciences examine the physical world. Physics, chemistry, biology, earth science, and astronomy all belong here.
- Social sciences examine how people think and how societies function. Psychology, sociology, economics, and political science are examples.
- Formal sciences focus on abstract systems and logic, like mathematics, computer science, and statistics. These fields provide the tools and frameworks that natural and social scientists use in their work.
The Science Council’s definition deliberately includes both the natural and social world, reflecting the fact that disciplines like psychology and economics use the same evidence-based methodology as chemistry or biology, even though their subjects are harder to isolate in a lab.
How Science Differs From Pseudoscience
One of the most influential ideas about what counts as science came from the philosopher Karl Popper in the early twentieth century. Popper argued that the key ingredient is falsifiability: a scientific claim must make specific predictions that could, in principle, be proven wrong by an experiment. If no possible observation could contradict an idea, that idea isn’t scientific.
Popper illustrated this with a comparison. Einstein’s theory of general relativity made precise predictions about how gravity would bend light. Those predictions could be tested, and if the measurements hadn’t matched, the theory would have been falsified. Freud’s theory of psychoanalysis, by contrast, could explain nearly any human behavior after the fact but made no specific predictions that an experiment could disprove. Popper considered the first scientific and the second not.
This distinction matters in everyday life. Pseudoscience, as the Stanford Encyclopedia of Philosophy defines it, is a collection of beliefs mistakenly regarded as being based on scientific method. The hallmarks are claims that can’t be tested, evidence that is anecdotal rather than systematic, and resistance to updating conclusions when contradictory data appears. Astrology, for instance, makes broad predictions that are reinterpreted to fit any outcome, which places it outside the boundaries of science regardless of how scientific its language sounds.
Why Peer Review Matters
Science is not just a method individuals follow. It’s also a collective institution with built-in checks. The most important of these is peer review, the process of subjecting a researcher’s work to scrutiny by other experts before it’s published. Reviewers evaluate whether the study’s design is sound, the data supports the conclusions, and the interpretation is reasonable. They can recommend that a paper be accepted, revised, or rejected entirely.
This system isn’t perfect, but it serves a critical function: it filters out unwarranted claims and personal opinions before they enter the published record. Because scientific knowledge is cumulative, with each new finding building on what came before, that filter protects the integrity of the entire structure. A single flawed study that slips through can mislead an entire line of research, which is why replication and ongoing scrutiny remain essential even after publication.