Plagiarism is one of the most common reasons a manuscript is delayed, rejected, or — in serious cases — retracted after publication. Yet most plagiarism in submitted manuscripts is not deliberate theft. It comes from sloppy note-taking, misunderstood paraphrasing, and citation habits formed under deadline pressure. The distinction matters for how you fix it, but it does not matter much for the outcome: a manuscript with substantial copied text will struggle regardless of intent.
This guide explains what counts as plagiarism, including the often-misunderstood case of self-plagiarism, how to paraphrase and cite properly, how similarity-check tools work and what their reports actually mean, and what to do if your paper is flagged. The aim is practical: habits that keep your work clean from the first draft, not just a definition.
Plagiarism is presenting someone else's work — words, ideas, data, images, or structure — as your own, without proper acknowledgement. Copying sentences word for word without quotation marks and a citation is the clearest case, but it is far from the only one. Using another author's distinctive argument, framework, or interpretation without credit is plagiarism even if every word has been changed. So is copying data, tables, or figures without permission and attribution, and so is translating another author's work into your language and presenting it as original.
Ideas need credit too, not just wording. If a paragraph in your paper follows another author's reasoning step by step, cite that author at the point where the reasoning appears, even if you wrote the sentences yourself. A citation at the end of a long passage does not cover the whole passage: place citations where the borrowed idea actually appears, so the reader can tell which thoughts are yours and which are borrowed.
Common knowledge is the main exception. Facts that are widely known in your field and not attributable to a particular author — for example, that a given disease affects a given species — do not need a citation for every mention. When in doubt, cite. An unnecessary citation costs nothing; a missing one can cost a manuscript.
Reusing your own previously published work without disclosure is also a problem, and it surprises many authors. It is often called self-plagiarism or text recycling. The issue is not ownership — the words are yours — but honesty to the reader and to the journal: a paper presented as new work should be new work. Copying large passages from your own earlier paper, submitting the same study to two journals, or publishing the same results twice without disclosure all breach publication ethics, even when no other author is involved.
Some reuse is legitimate. Describing the same method in a similar way across papers from one project is normal, as is building on your earlier findings with a citation to them. The line is crossed when the overlap is substantial and undisclosed: a reader who reads both papers should not discover that the second is largely the first in different formatting. When you reuse material from your own earlier work, cite it like any other source, and tell the editor in your cover letter if the overlap is more than incidental.
Thesis work deserves a note here. A thesis chapter can often become a journal article, but it usually needs substantial revision: a journal article is shorter, more focused, and written for readers who have not read the thesis. Check the journal's policy on prior work, disclose the thesis in your cover letter if the journal asks for it, and make sure the article stands on its own. Our guide on how to publish a research paper covers this in more detail.
Proper paraphrasing means restating an idea in your own words and structure, with a citation. It is not a word-swap exercise, and it is not a matter of changing every third word while keeping the sentence shape. Editors and similarity tools both recognise that pattern easily.
Read the source, close it, and write the idea from your understanding. If you cannot write it without looking, you have not yet understood it well enough to paraphrase it — read it again, think about what it actually claims, and then write. A good paraphrase usually changes the sentence structure, not just the vocabulary: it may combine two sentences, split a long one, or reorder the logic. It also fits the argument of your own paper, rather than sitting in it as an imported block. And it carries a citation, because paraphrasing without a citation is still plagiarism.
Patchwriting — stitching together lightly altered phrases from a source — is the most common form of unintentional plagiarism, especially among authors writing in a second language. It happens when you write with the source open in front of you and lean on its phrasing. The defence is procedural: take notes in your own words from the start, never copy passages into your notes without marking them as quotations, and draft with your notes, not with the sources, open. If a sentence in your draft sounds suspiciously polished compared with the sentences around it, check it against the source; it may be closer to the source than you realised.
Direct quotation is sometimes the right choice: for a definition, a particularly precise formulation, or a statement you are analysing. When you quote, use quotation marks, cite the source, and keep the quotation short. Long block quotations have little place in a research article; if you need more than a sentence or two of someone else's words, you probably need to paraphrase instead, or to reconsider whether the passage belongs in your paper at all.
Most plagiarism problems disappear when citation becomes a habit rather than an afterthought. Cite as you write, not at the end: every time you use an idea, a result, or a phrase from a source, put the citation in immediately. Retrospective citation — going back through a finished draft trying to remember where each idea came from — is unreliable, and the ideas you cannot trace are the ones most likely to be flagged.
Use a reference manager from the start of a project, and check every reference against the original source rather than copying a citation from another paper. Errors in volume, page, or year propagate easily through copied citations, and reviewers notice them. Make sure every in-text citation appears in the reference list and every reference is cited in the text; a mismatch in either direction is a screening flag.
Follow the journal's reference style exactly. Some journals number references in order of appearance; others use author and year in the text. Set your reference manager to the journal's style at the start of writing, because switching styles late is slow and error-prone. See the animal sciences manuscript guidelines for field-specific citation and formatting expectations.
Most journals run submitted manuscripts through a similarity-check tool — widely used examples include Turnitin and iThenticate — which compare the text against a large database of published work, web pages, and previously submitted manuscripts. The tool produces a similarity report showing matched passages and their sources, along with an overall similarity percentage.
That percentage is not a plagiarism verdict. A report can show a high percentage for innocent reasons: standard method descriptions, commonly used phrases, reference lists, and properly quoted material all match existing text. Conversely, a low percentage does not prove originality: a paraphrased passage, a translated passage, or an idea taken without its wording may match nothing at all. Editors read the report, not just the number — they look at which passages matched, how long they are, and whether they are cited.
If your institution gives you access to a similarity tool, use it on your draft before submission, and read the report the way an editor would: check each matched passage, not the total. Fix genuine matches by paraphrasing properly or quoting with citation, and leave alone the matches that are standard phrases, references, or your own properly cited method descriptions. If you do not have access to a tool, the habits in this guide — own-word notes, cite-as-you-write, and a final source check — are the same habits the tools reward.
Be cautious with free online checkers that ask you to upload a full manuscript. Some retain uploaded text, which can create problems if the same text is later checked by a journal's tool. Prefer your institution's tool or the journal's own process, and never upload unpublished work to a service whose terms you have not read.
If an editor tells you the similarity report needs attention, do not panic and do not argue about the percentage. Ask which passages are a concern, read them against their sources, and fix each one: rewrite it properly in your own words with a citation, quote it correctly, or remove it if it is not needed. Then tell the editor what you changed, passage by passage, in your response. Editors deal with flagged manuscripts routinely, and a prompt, thorough correction usually resolves the matter at the revision stage.
If the flag concerns your own earlier work, disclose the overlap, cite the earlier paper, and explain the relationship between the two papers in your response letter. Transparency is the remedy for text recycling in the same way that citation is the remedy for borrowed wording: the problem is undisclosed overlap, not the existence of related work.
If you believe a flag is mistaken — for example, a matched passage that is a standard method description or a properly cited quotation — say so calmly, with the evidence: point to the citation, the quotation marks, or the standard phrasing. Editors can distinguish a mistaken flag from a real problem, but they need you to show them, passage by passage, rather than dismissing the report as a whole.
Serious cases — extensive copied text, or a paper submitted while the same work is under consideration elsewhere — can lead to rejection and, after publication, to retraction and notification of the author's institution. That is why the habits in this guide matter before submission, not after: prevention is a matter of note-taking and citation discipline, while a cure after publication is a formal process with lasting consequences.
Originality in research writing is not about never using other people's work — scholarship is built on it. It is about making the boundary between your contribution and everyone else's visible, so that readers, reviewers, and editors can see exactly what you add. That boundary is created by two simple disciplines: your own words, and visible credit. Keep both, and plagiarism stops being a risk to your paper and becomes a non-issue.
| Situation | Is it plagiarism? | What to do instead |
|---|---|---|
| Copying sentences word for word without quotation marks | Yes | Quote with marks and a citation, or paraphrase properly |
| Restating an idea in your own words with a citation | No | This is correct paraphrasing — keep doing it |
| Restating an idea in your own words without a citation | Yes | Add the citation at the point the idea appears |
| Swapping a few words while keeping the sentence structure | Yes — patchwriting | Close the source and rewrite from your understanding |
| Reusing large passages from your own earlier paper without disclosure | Yes — self-plagiarism | Cite the earlier work and disclose the overlap to the editor |
| Using widely known facts in your field without a citation | No — common knowledge | When in doubt, cite anyway; it costs nothing |
| Copying a table, figure, or dataset without permission and credit | Yes | Seek permission where required and credit the source |
These authoritative external resources explain the standards and guidance mentioned in this article:
Is paraphrasing without a citation still plagiarism?
Yes. Changing the words does not remove the need for credit. A paraphrase must carry a citation at the point where the borrowed idea appears.
Can I reuse text from my own published paper?
Only in small amounts and with disclosure. Substantial undisclosed reuse of your own earlier work counts as self-plagiarism; cite the earlier paper and tell the editor about the overlap.
What similarity percentage is acceptable?
There is no universal safe percentage. Editors read the report passage by passage: standard phrases, references, and properly cited material are fine, while long unmatched or uncited passages are a problem whatever the total.
Do I need to cite common knowledge?
No, facts widely known in your field do not need a citation for every mention. When you are unsure whether something counts as common knowledge, cite it — an unnecessary citation costs nothing.
Is translating another author's work into my language plagiarism?
Yes, if presented as your own without credit. Translation does not change authorship; cite the original source.
What should I do if my paper is flagged for similarity?
Read the flagged passages against their sources, fix each one by proper paraphrasing, quotation, or removal, and tell the editor passage by passage what you changed.