Science vs. Pseudoscience: How to Tell the Difference
Distinguishing science from pseudoscience is one of the most practically important skills a person can develop — and one of the least formally taught. This page examines the defining features of each, the mechanisms that separate them, the real-world contexts where the boundary gets blurry, and the specific markers that reliably signal which side of the line a claim falls on.
Definition and scope
Karl Popper, the 20th-century philosopher of science, identified falsifiability as the central criterion separating science from non-science. A scientific claim must be structured in a way that allows it to be proven wrong. "Vaccines cause autism" is a falsifiable claim — and it has been tested and refuted across studies involving millions of children, according to the CDC. "Crystals realign your spiritual energy" is not falsifiable; no experiment could ever demonstrate it to be wrong, which places it outside science entirely.
Pseudoscience mimics the surface features of science — it uses technical-sounding language, cites statistics, and sometimes references real studies — but it systematically avoids the mechanisms that make science self-correcting. The National Center for Science Education defines pseudoscience as claims presented as scientific while failing to meet scientific standards of evidence, methodology, and peer scrutiny.
The scope of this distinction matters well beyond academic philosophy. The US Federal Trade Commission has taken enforcement action against health product companies making unsupported scientific claims, and the National Institutes of Health maintains research offices specifically to evaluate claims that exist in the contested zone between alternative medicine and evidence-based practice.
How it works
Science operates through a specific set of interlocking mechanisms — none of which is optional. Understanding how those mechanisms function is the subject of How Science Works, but for the purpose of comparison, the core difference between science and pseudoscience can be broken into five structural features:
- Falsifiability — Scientific hypotheses are framed so that a specific result would disprove them. Pseudoscientific claims shift or expand to accommodate contradictory evidence rather than being revised by it.
- Peer review and replication — Scientific findings are submitted to independent scrutiny and must be reproducible. Pseudoscientific claims frequently rely on a single study, unpublished data, or testimonials.
- Error correction — Science has a built-in mechanism for abandoning wrong ideas when evidence accumulates against them. Pseudoscience treats criticism as persecution and disconfirming evidence as a conspiracy.
- Proportionality of evidence — Extraordinary claims require extraordinary evidence (Carl Sagan, Cosmos, 1980). Pseudoscience inverts this, treating the absence of disproof as proof.
- Transparent methodology — Scientific studies describe their methods in enough detail that others can replicate or challenge them. Pseudoscientific claims often rely on proprietary formulas, secret processes, or the credibility of a single charismatic source.
The gap between these two systems is not subtle. It is structural.
Common scenarios
The science-pseudoscience boundary appears with surprising frequency in everyday decisions.
Health and medicine is the highest-stakes arena. Homeopathy claims that water retains a "memory" of dissolved substances at dilutions so extreme that not a single molecule of the original compound remains. The proposed mechanism contradicts basic chemistry and has not survived rigorous testing. A 2015 systematic review by Australia's National Health and Medical Research Council examined 176 homeopathy trials and found no good-quality evidence that homeopathy was more effective than placebo for any health condition (NHMRC, 2015).
Nutrition science is a domain where genuine scientific uncertainty gets weaponized. The replication crisis in nutritional epidemiology is real — the Stanford Prevention Research Center's John Ioannidis has published extensively on the methodological problems in food and health research. Legitimate debate about dietary fat, for instance, differs fundamentally from claims that a single supplement cures cancer, because the former involves contested evidence and the latter suppresses or ignores disconfirming evidence.
Psychology has its own contested zone. Facilitated communication — a technique claiming to allow non-verbal individuals to express thoughts through a facilitator's hand-guided typing — was treated as a breakthrough in the 1990s before controlled studies demonstrated that the facilitator, not the patient, was generating the communication. The American Psychological Association issued a formal position statement against it.
Decision boundaries
When evaluating an unfamiliar claim, four questions reliably separate scientific from pseudoscientific territory:
Does the evidence come from controlled studies, or from testimonials? Testimonials are the currency of pseudoscience. A single dramatic success story, stripped of base rates and confounding factors, proves nothing about efficacy.
Has the claim survived hostile peer review? Not friendly journals with low standards, not self-published monographs, but scrutiny from researchers specifically trying to find flaws.
What would falsify this claim? If a proponent cannot name a specific result that would convince them they are wrong, the claim is not operating scientifically.
Does the claim require a separate explanation for why it hasn't been accepted? When a theory's proponents argue that establishment science suppresses the truth, that the mainstream is corrupt, or that only outsiders see clearly — that pattern is a reliable signal of pseudoscience, not a sign of revolutionary thinking.
The Science home page contextualizes these principles within the broader question of how scientific knowledge is built and evaluated. Crossing the boundary from pseudoscience to science requires more than better marketing — it requires submitting to the mechanisms that make science capable of being wrong, and therefore capable of being right.