Yes, plagiarism can apply to math formulas, but only to the parts that carry authorship: the specific notation, the worked derivation, the surrounding explanation, and the selection and arrangement of symbols. The underlying mathematical truth—that a formula correctly relates quantities—is not ownable and cannot be plagiarized. What gets judged is how a person expresses, derives, and presents that truth, not the truth itself.
This distinction matters because math sits between two extremes. Prose is fully copyrightable expression. A law of nature or a basic fact is not. Formulas live in the contested middle, and the answer depends on which layer of a formula you are looking at. A copied equation with different variable names is usually fine as mathematics. A copied derivation with the same steps, the same intermediate expressions, and the same idiosyncratic notation is a different matter.
The cleanest way to reason about it is to separate the idea from the expression and ask which one traveled. Math has a thin copyright doctrine, meaning few ways exist to write the same true relationship, so courts and institutions tolerate heavy similarity at the level of bare formulas. Authorship enters through derivation, exposition, notation choices, and compilation. Plagiarism is a broader and stricter standard than copyright infringement, and academic integrity rules often punish copying that a court would never touch.
Mechanism: Why Formulas Escape and Where They Do Not
Mathematics is treated as a system of facts and methods, not a literary work. Under United States copyright doctrine, ideas, procedures, processes, systems, and methods of operation are excluded from protection. A formula that states a true relationship is closest to a fact or a method, so the bare equation is generally free to reuse. This is the doctrinal reason a textbook cannot stop you from writing the quadratic formula.
Copyright protects the expression, and in math the expression is thin. When only a handful of ways exist to write a correct equation, the merger doctrine limits protection: if idea and expression are inseparable, copying the expression is not infringement. That is why variable renames and trivial reordering usually defeat a claim at the level of the equation itself. The formula is the idea, and the idea is not owned.
Authorship re-enters through the surrounding apparatus. A multi-step derivation with a particular sequence of substitutions, a distinctive choice of symbols, an annotated proof, a table of worked examples, or a chapter's organization of formulas all carry creative selection. Copying that apparatus without attribution is where plagiarism findings actually arise. The equation is free; the journey to it and the packaging around it are not.
One level deeper: attribution norms in mathematics are about credit for discovery and credit for exposition, and they run on different tracks. The discoverer of a theorem gets priority credit by convention, which is why names attach to results. The expositor gets credit for a clear derivation or a useful formulation. Plagiarism rules mainly police the second track, because that is where copying is detectable and where the copied work is genuinely someone's writing rather than a discoverable truth.
Quantified Context: How Often and How Much
There is no authoritative body that publishes a plagiarism-detection rate specific to math formulas, and inventing a percentage would misrepresent the evidence. What can be stated is the shape of the problem. Similarity-detection tools such as Turnitin compare text strings, so they handle prose well and equations poorly. Formulas embedded as images or as typeset math often pass through text-matching systems without a flag, which means detection depends heavily on whether the equation was written as searchable text.
Institutional definitions of plagiarism are broader than detection. Most university academic-integrity policies cover copying ideas, data, and "work" without attribution, not just verbatim words, so a copied derivation can violate policy even when no software catches it. The threshold that matters is the policy's wording, not a similarity score. A common practical threshold in policy text is any unattributed use of another person's specific expression or substantial selection, which is a qualitative test rather than a number.
The asymmetry is the counterintuitive detail. A student who copies a block of prose is likely to be flagged. A student who copies an entire derivation, retypes it with different variable letters, and presents it as original may pass every automated check while committing the clearer violation. Detection and wrongdoing are only loosely correlated in mathematics.
Edge Cases: Where the Line Actually Sits
Standard formulas that appear in every reference—the Pythagorean theorem, the quadratic formula, the normal distribution density—are common knowledge and need no citation. They are facts of the field. Attribution here would be noise, and no reasonable reviewer treats writing them as plagiarism.
Named theorems and results are a middle case. You do not need to cite the equation itself, but you should credit the originator by name, as convention already does: Bayes' theorem, Euler's formula, the Cauchy-Schwarz inequality. The citation is to the result's provenance, not permission to use the math.
A copied derivation is the clearest violation. If you lift the specific sequence of steps, the intermediate algebra, and the explanatory asides that get from a problem to a formula, you have copied expression. Renaming variables does not launder it, because the creative content is the path and the arrangement, not the letters.
A distinctive notation scheme is expression. If an author invents a particular symbol system, a custom operator, or an unusual layout for a family of equations, copying that scheme wholesale is copying a creative choice. Standard notation carries no authorship; idiosyncratic notation can.
Figures and diagrams built from formulas deserve separate care. A plot, a geometric construction, or a formatted equation image may be protected as a visual work even when the underlying math is free, and reusing the image without permission or credit is a distinct problem from reusing the equation.
Red-flag warnings:
Retyping a derivation with renamed variables and calling it your own. Pasting an equation image lifted from a paper without credit. Reproducing another author's idiosyncratic notation as if it were standard. Presenting a compiled table of formulas and their worked examples as original organization. Omitting a citation for a specialized result that is not common knowledge in your audience.
Misconceptions: What People Get Wrong
The most common error is treating math as exempt because "facts can't be copyrighted." That is half right. The fact is free, but the derivation, the exposition, and the arrangement are not, and plagiarism rules reach further than copyright. Exemption at one layer is mistaken for exemption at every layer.
The second error is believing that changing symbols defeats plagiarism. Variable renaming addresses the equation, which was never the protected part. It does nothing to the copied derivation, which is the part that carries authorship. The trick protects against the wrong charge.
The third is conflating plagiarism with copyright infringement. Plagiarism is an attribution and honesty norm enforced by institutions, while copyright is a legal claim enforced by courts. You can plagiarize material that is out of copyright, and you can infringe without plagiarizing if you have permission but no attribution requirement. Math sits where the two standards diverge most.
The fourth is over-citing to the point of absurdity, crediting Euclid for the Pythagorean theorem in every paragraph. Attribution norms track what a knowledgeable reader would consider common knowledge. Over-citation of facts signals uncertainty rather than rigor, so the goal is calibrated credit for genuine authorship, not ceremonial citation of the whole canon.
Practical Actions: How to Stay Clean
The working rule is to cite expression and credit discovery, while freely using facts. Apply that rule step by step.
Use standard formulas without citation when they are common knowledge for your audience. Cite the source when you reproduce a specific derivation, a distinctive notation, or a compiled presentation. Rewrite derivations in your own sequence where the math allows, and cite the original approach if you follow it closely. Credit named results by their conventional names instead of presenting them as your own findings. Treat figures and formatted equation images as separate works that need permission or credit. Check your institution's integrity policy wording, because its threshold governs findings regardless of what detection software reports.
For a threshold that reconciles the competing standards: if a reader with field knowledge would recognize a specific passage as someone else's distinctive work, attribute it. If the content is a bare true relationship any competent writer would produce identically, do not. When a case falls between those poles—a somewhat standard derivation with a particular flavor—cite, because the cost of a needless citation is far lower than the cost of an integrity finding.
Reconciling the thresholds explicitly: copyright asks whether expression was copied, plagiarism policy asks whether unattributed work was presented as your own, and detection tools ask whether text strings match. These three can disagree, and the binding standard for a student or author is usually the institutional policy, which is stricter than copyright and broader than detection. Treat the strictest applicable standard as the one that governs your conduct.
Limitations: What This Cannot Resolve
This analysis cannot tell you how a specific institution will rule on a specific submission, because policies vary and cases are judged on context. It also cannot substitute for legal advice on copyright, which turns on jurisdiction and specific facts. The stable principles are clear; the application to a single document is not mechanical.
There is also a genuine gray zone that no rule eliminates. Some derivations are so standard that independent authors produce near-identical versions, and distinguishing independent convergence from copying requires judgment no threshold captures. Mathematics has few ways to write a true result, which is exactly why expression is thin and why honest authors sometimes match without plagiarizing. The professional caveat is to document your process—notes, drafts, and sources—so that similarity can be explained by method rather than assumed to be copying.
The practical takeaway holds across these limits. The formula is not the thing to worry about; the derivation, notation, exposition, and compilation are. Attribute those, use facts freely, and let the math itself remain common property.