03 · What You Need to Know
Plagiarism Is Broader Than Copying Someone Else's Sentences
A useful definition starts with appropriation and credit
The U.S. Office of Research Integrity (ORI) defines plagiarism for purposes of its research-misconduct framework as the appropriation of another person's "ideas, processes, results, or words without giving appropriate credit." That definition is useful because it immediately moves the issue beyond a simplistic copy-and-paste test.
The object being appropriated may be language, but it may also be an intellectual contribution that survives even after the language has changed. The practical question is therefore not merely, "Did I copy these words?" It is also, "Whose intellectual contribution am I presenting here, and have I made its origin sufficiently clear?"
Verbatim copying can be plagiarism
The clearest case is taking text from another source and presenting it as though you wrote it. ORI's guidance on plagiarism of text describes this as copying text from another source without appropriate credit. When another author's wording is reproduced verbatim, scholarly conventions generally require the borrowing to be identified as a quotation, together with the appropriate citation.
A reference alone does not necessarily turn copied wording into a proper paraphrase. If you reproduce distinctive wording while making it appear that the prose is yours, the reader may still be misled about who wrote those words. The exact conventions for quotations, page numbers, and citation format vary by discipline and style.
Changing a few words may still leave the borrowing intact
Replacing selected words with synonyms, rearranging a sentence, or stitching together lightly modified phrases does not necessarily create independent prose. ORI discusses forms of patchwriting or mosaic plagiarism in which borrowed language and ideas remain closely tied to the original even though portions have been altered.
A genuine paraphrase requires you to express the source material in your own formulation while preserving its meaning accurately. More importantly, paraphrasing does not make the underlying contribution yours. When the idea, finding, interpretation, or other attributable material came from a source, you ordinarily still need to cite the source even after paraphrasing it.
Ideas can be plagiarized without copying their wording
Suppose you encounter a novel hypothesis, conceptual explanation, interpretation, or theoretical argument in another researcher's work. You completely rewrite it and use none of the author's original sentences. The absence of copied wording does not automatically solve the attribution problem.
ORI specifically includes ideas within its definition of plagiarism and describes idea plagiarism as appropriating another person's idea, in whole or in part or with superficial modification, without giving appropriate credit. This is why plagiarism can occur without copying the original words.
Results and data-derived contributions also require attribution
Research findings do not become yours simply because you describe them differently. If another researcher produced a result and you report or use that result in a way that implies it arose from your own work, the problem concerns the origin of the research contribution, not merely its wording.
Researchers should be particularly careful when moving information among literature notes, datasets, draft manuscripts, collaborative documents, and their own prose. Once the original source disappears from your notes, borrowed material can become surprisingly easy to mistake for something you produced yourself.
Processes and methods can raise attribution questions too
ORI's formal definition also includes "processes." This does not mean that every use of a standard statistical test, laboratory procedure, survey technique, or established research design is plagiarism. Scholarly methods are routinely reused, and what requires attribution depends partly on whether the procedure is standard knowledge or represents an identifiable contribution from another source.
A distinctive procedure, adaptation, protocol, or methodological innovation derived from another researcher may require acknowledgment even if you independently describe how you implemented it. The more specific question of when this boundary is crossed is better considered in the context of whether you can plagiarize a research method.
Figures, tables, frameworks, and conceptual models are not exempt
Borrowing is not limited to prose. A researcher's figure, table, framework, diagram, or conceptual model may embody intellectual or expressive work whose origin should be acknowledged. Recreating an item rather than taking a screenshot does not automatically make it original.
The appropriate treatment can also involve a separate copyright question. Depending on what is reused, how much is reproduced, the applicable law, the license, and publisher policy, citation may not be the only consideration. The detailed rules for figures, tables, frameworks, and conceptual models therefore deserve separate treatment.
Plagiarism
Concerns whether another person's intellectual contribution is represented without appropriate credit.
Copyright infringement
Concerns legal rights governing protected expression and its reproduction, distribution, adaptation, or other uses.
Plagiarism and copyright infringement are not the same thing
The distinction matters. Under U.S. copyright law, copyright can protect original expression but does not protect underlying facts, ideas, procedures, processes, systems, methods of operation, concepts, principles, or discoveries. Scholarly attribution norms can nevertheless require you to acknowledge the source of an idea or discovery.
Consequently, conduct can raise a plagiarism problem without necessarily constituting copyright infringement. Conversely, giving a citation does not by itself establish that you have legal permission to reproduce copyrighted material. Copyright rules vary by jurisdiction, and licensing and publisher agreements may add further conditions.
Plagiarism can occur before publication
Plagiarism is not restricted to the final journal article. Under the U.S. federal research-misconduct framework, plagiarism may occur in proposing, performing, reviewing, or reporting research. The federal policy also addresses the misuse of material encountered during confidential peer review.
That means research proposals, theses, dissertations, conference materials, peer-review activities, and other parts of the research record can raise attribution concerns. Confidential material presents an additional issue: citing something obtained through confidential review does not necessarily give you permission to use it.
Plagiarism is not determined by a universal similarity percentage
There is no general rule that a manuscript becomes plagiarism at a particular similarity percentage. Text-matching software identifies textual overlap. It does not independently determine why the overlap exists, whether it has been appropriately quoted, whether the matched language is routine terminology, or whether plagiarism of an idea occurred without textual similarity.
COPE discussions of text recycling likewise caution against treating percentages as decisive because the nature and location of overlap can matter more than the raw number. A small but intellectually consequential appropriation may matter greatly, while a larger amount of legitimate matching material may require a very different interpretation.
Watch Out
Do not treat a similarity score as a plagiarism verdict. A similarity report is evidence to inspect, not a substitute for examining the matched passages, their sources, attribution, context, and scholarly purpose.
Plagiarism is not identical to a formal finding of research misconduct
Another distinction is easily lost. A questionable passage may warrant correction or raise a plagiarism concern without automatically satisfying every requirement for a formal institutional or regulatory finding of research misconduct.
For example, under the U.S. federal framework described by ORI, a formal finding of research misconduct requires a significant departure from accepted practices of the relevant research community, conduct committed intentionally, knowingly, or recklessly, and proof by a preponderance of the evidence. Honest error and differences of opinion are excluded from the federal definition of research misconduct.
This distinction is particularly relevant to accidental plagiarism. Lack of intent does not make an attribution problem disappear, but intent can matter considerably when institutions determine whether conduct meets a formal definition of misconduct and what response is appropriate.
07 · A Quick Checklist
Check Your Draft for Attribution Problems Before Submission
Before submitting your research, check:
Can I identify the source of every quotation, close paraphrase, distinctive idea, and borrowed finding in my draft?
Have I clearly marked wording that I reproduced verbatim rather than making it look like my own prose?
Have I cited sources for substantive ideas and findings even when I expressed them entirely in my own words?
Have I avoided patchwriting that merely substitutes, deletes, or rearranges words from the source?
Have I acknowledged distinctive methods, frameworks, models, figures, tables, or other intellectual contributions that I borrowed or adapted?
If I used a similarity checker, have I examined the actual matches instead of treating the percentage as a plagiarism verdict?
Have I traced any passages whose origins became unclear while moving material between notes and drafts?
Where substantial material is reproduced or adapted, have I checked whether copyright, licensing, or permission requirements apply in addition to citation?
