Close paraphrasing is the superficial modification of material from another source. Editors should generally summarize source material in their own words, adding inline citations as required by the sourcing policy.

Limited close paraphrasing is appropriate within reason, as is quoting (with or without quotation marks), so long as the material is clearly attributed in the text – for example, by adding "John Smith wrote ...," together with a footnote containing the citation at the end of the clause, sentence or paragraph.

Close paraphrasing without in-text attribution may constitute plagiarism, and when extensive (with or without in-text attribution) may also violate Wikipedia's copyright policy, which forbids Wikipedia contributors from copying material directly from other sources. Public domain material must likewise be attributed to avoid plagiarism. If the source material bears a free copyright license that is compatible with Wikipedia's licenses, copying or closely paraphrasing it is not a copyright violation so long as the source is attributed somewhere in the article, usually at the end.

The best way to prevent close paraphrasing is to understand clearly when it is a problem, how to avoid it, and how to address it when it appears.

There are legal, ethical, and organizational standard considerations regarding the use of close paraphrasing.

Copyright law

Wikipedia's primary concern is with the legal constraints imposed by copyright law. Close paraphrasing of the creative expression in a non-free copyrighted source is likely to be an infringement of the copyright of the source. In many countries close paraphrasing may be also seen as mutilation or distortion of an author's work, infringing on their moral rights.

Creative expression

Facts and ideas cannot be protected by copyright, but creative expression is protected.

Hilaire Belloc illustrates two forms of creative expression in his 1897 More Beasts: (for Worse Children), where he informs us that "The Llama is a woolly sort of fleecy hairy goat, with an indolent expression and an undulating throat; like an unsuccessful literary man". If this somewhat dubious source was used for the article on Llama and was still protected by copyright, it would be acceptable to say that the Llama is an animal with a shaggy coat, and perhaps that it has a long neck. These are facts. But use of the phrases "indolent expression" and "undulating throat" would violate copyright. The original choice of words is part of Belloc's creative expression. Going further, the simile "like an unsuccessful literary man" is also creative, and is also protected. A clumsy paraphrase like "resembling a failed writer" would violate copyright even though the words are entirely different. More than the facts have been copied.

Although facts are not subject to copyright, a selection or arrangement of facts may be considered creative and therefore protected. For example, an alphabetical list of states in the USA giving their name, size and population cannot be copyrighted. However, a shorter list of states giving the name, size and population as before, but ranked as the "top most livable states" would be subject to copyright. The selection and ranking is creative. To avoid problems, editors should avoid any fanciful wording, imaginative metaphors or similes, or subjective interpretations, and stick to giving the facts in plain words. That is what the reader wants from an encyclopedia.

Moral rights

The "moral rights" of an author are independent of copyright ownership. They include the author's right to control first publication of a work; the author's right to be attributed or to remain anonymous; the author's right for the work to be published without distortion or mutilation. As with copyright, moral rights apply to creative expression but not to mere facts. Editors must respect moral rights to ensure that Wikipedia content can be reused as widely as possible.

Wikipedia editors must not use unpublished work in any way. With published work, editors should attribute each source to the author where the publication names the author, and attribute the source to the publication if it does not name the author. It is sometimes relevant for an article to include a short quotation such as a significant statement made by the subject of the article or a notable comment about the subject. In these cases a verbatim quotation should be given rather than a paraphrase. Quotations should be used very sparingly and must be clearly identified and formatted as defined in MOS:QUOTE.

Substantial similarity

Depending on the context and extent of the paraphrasing, limited close paraphrase may be permitted under the doctrine of fair use. An author may think they are being original when they write "Charles de Gaulle was a towering statesman", not realizing that many other authors have independently come up with these identical words. Copying or close paraphrasing may thus be accidental. Two reasonable people may disagree on whether a paraphrase of one sentence copies creative expression or only reproduces facts, which are not protected. There is thus an element of subjectivity. The concept of "substantial similarity" is used to weed out these trivial instances.

Paraphrasing rises to the level of copyright infringement when there is substantial similarity between an article and a source. This may exist when the creative expression in an important passage of the source has been closely paraphrased, even if it is a small portion of the source, or when paraphrasing is looser but covers a larger part of the source. A close paraphrase of one sentence from a book may be of low concern while a close paraphrase of one paragraph of a two-paragraph article would be considered a serious violation. Editors must therefore take particular care when writing an article, or a section of an article, that has much the same scope as a single source. The editor must be extra careful in these cases to extract the facts alone and present the facts in plain language, without carrying forward anything that could be considered "creative expression".

Wikipedia's guidelines

Even when content is verifiably public domain or released under a compatible free license (see Wikipedia:Donating copyrighted materials), close paraphrasing may be at odds with Wikipedia's guideline related to plagiarism (see Wikipedia:Plagiarism). While in this context, too, close paraphrasing of a single sentence is not as much of a concern, if a contributor closely paraphrases public domain or freely licensed content, he or she should explicitly acknowledge that content is closely paraphrased. (See above.)

Another potential problem arises when a contributor copies or closely paraphrases a biased source either purposefully or without understanding the bias. This can make the article appear to directly espouse the bias of the source, which violates our neutral point of view policy.

There are a few specific situations when close paraphrasing is permitted. If information is gathered from the public domain or is free use content, close paraphrase may be acceptable. In some instances it is helpful to capture the words as written, in which case the guidelines for Quotations apply. Lastly, there may be some instances where it's difficult to paraphrase because of the nature of the content, in such cases there are a couple of tips about how to limit the degree of close paraphrasing.

When using a close paraphrase legitimately, citing a source is in most cases required and always highly recommended.

Public domain or free use content

In some limited cases, close paraphrase may be an acceptable way of writing an article. For example, many Wikipedia articles are (or were) based on text from the 1911 Encyclopedia Britannica (see Wikipedia:1911 Encyclopaedia Britannica). If the source is public domain, such as work of the U.S. government, or available under a CC BY-SA-compatible free license, it may be closely paraphrased if it is fully attributed. Acknowledging the source in such instances may include accompaniment by in-text attribution that makes clear whose words or ideas are being used (e.g. "John Smith wrote that ...") or may include more general attribution that indicates the material originates from a free source, either as part of an inline citation or as a general notice in the article's "References" section (see Wikipedia:Plagiarism#Copying material from free sources).

Quotation of non-free text

Limited quotation from non-free copyrighted sources is allowed, as discussed in Wikipedia's non-free content policy and guideline. Quotations should have in-text attribution and should be cited to their original source or author (see WP:When to cite). With direct quotation, editors should clearly distinguish the quoted material from the original text of the article following the guidelines for quotations. Extensive use of quotation from non-free sources is generally not acceptable. Even if content is attributed, it can still create copyright problems if the taking is too substantial. To avoid this risk, Wikipedia keeps this—like other non-free content—minimal.

Quotation from non-free sources may be appropriate when the exact words in the source are relevant to the article, not just the facts or ideas given by the source. Examples may include statements made by a person discussed in the article, brief excerpts from a book described in the article, or significant opinions about the subject of the article. Quotation should not, however, be treated as an alternative to extracting facts and presenting them in plain language. Thus:

• Right: Churchill said, "I have nothing to offer but blood, toil, tears and sweat."^[1]
• Right: The New York Times reviewer found the film "pretentious and boring".^[2]
• Wrong: According to Bulgarian Butterflies, "the patient observer may be fortunate enough to glimpse this rare moth flitting along the mossy banks of a woodland stream."^[3]

When there are a limited number of ways to say the same thing

Close paraphrasing is also permitted when there are only a limited number of ways to say the same thing. This may be the case when there is no reasonable way to avoid using a technical term, and may also be the case with simple statements of fact.

It is acceptable to use a technical term such as "The War of the Spanish Succession" or "Relational Database Management System (RDBMS)" when the term is almost always used by sources that discuss the subject, and when such sources rarely use any other term. In this case, the technical term is considered to be "merged" with the idea expressed. There is no reasonable alternative way of expressing the idea, and since ideas are not subject to copyright the term is also not protected. However, if different sources use different terms for the concept, it is best for the article to use a different term from the source.

An example of closely paraphrased simple statements of fact is given by a biography that relies on two sources for the basic outline. The sources and the article start with:

• Source1: John Smith was born in Hartford, Connecticut on February 2nd 1949... He attended State University, obtaining an M.D. in 1973.
• Source2: John Smith was born on 2 February 1949 in Hartford... He graduated with a medical degree from State University in 1973.
• Article: John Smith was born on 2 February 1949 in Hartford, Connecticut... He studied medicine at State University, and earned an M.D. in 1973.

In this example, the wording of the article is very close to that of both sources. However, the article merely presents standard facts for a topic like this in standard sequence. The article does not copy any creative words or phrases, similes or metaphors, and makes an effort at paraphrasing in the second sentence. Just two short sentences are close to the sources. For these reasons the close paraphrasing should be acceptable. Note, however, that closely paraphrasing extensively from a non-free source may be a copyright problem, even if it is difficult to find different means of expression. The more extensively we rely on this exception, the more likely we are to run afoul of compilation protection.^[1]

In this example, Wikipedia's article text is an attempt at paraphrasing the source. However, almost all of the original word choice, word order and sentence structure is retained.

Analysis:

• "A statement issued by the receiver, Deloitte's David Carson, confirmed that, of the 670 employees, 480 of them would be laid off" vs. "A statement from the receiver, David Carson of Deloitte, confirmed that 480 of the 670 employees have been made redundant".
  • The structure of Wikipedia's statement is essentially the same as the original. Changing a single word and slightly reordering one phrase is not enough to constitute a paraphrase.
• "They insisted they would refuse to leave until they had met with Carson" vs. "The employees say they will not be leaving until they meet with Mr. Carson".
  • The structure of this sentence is the same as the original with too much similarity within the structure of the paragraph.
• "there was a minor scuffle during which the main door to the visitors' centre was damaged" vs. "There were some scuffles at one point and a main door to the visitors' centre was damaged".
  • The structure and language of the two sentences are the same.
• "Local Sinn Féin Councillor Joe Kelly was amongst those who occupied the visitors' gallery" vs. "Local Sinn Féin Councillor Joe Kelly, who is one of those currently occupying the visitors' gallery".
  • This slight rewording does not change the fact that the underlying structure and language are the same. Minor changes, such as "was amongst those" --> "is one of those" and "occupied" --> "currently occupying", are not enough to constitute an original rewriting of the passage.

See below for an example of an unusable paraphrase repaired to become acceptable.

To properly paraphrase content, you review information from reliable sources, extract the salient points, and use your own words, style and sentence structure to draft text for an article.^[4][5]

Take notes

One of the key factors in the creation of inadvertent close paraphrasing is starting with text taken directly from the source. The word choice and style can easily resurface since it is foremost in our minds. An approach to ensure that information is fully understood and formed into unique thoughts is to isolate the essential information by taking notes.

• Start off by taking notes of essential information only, excluding the use of phrases.
• Record the source for the citation.
• Collect information from multiple sources. This will have several benefits: it promotes a tone with a neutral point of view and produces a well-rounded understanding of the topic. It also makes it less likely that your end result will follow too closely on any one source.
• Gather short quotations when they powerfully illustrate a point for your article. Overuse can result in a disjointed article and may breach copyright. (Extensive quotations are forbidden by policy.)

Write a draft

• Gather related items from the multiple sources and explain it to yourself: The point is to rephrase or summarize a body of information in your own words and sentence structure.
• Add inline citations in accordance with the sourcing guideline.
• Allow time between note-taking and drafting to clear your mind of the original diction and better paraphrase the content.
• Don't paraphrase information in the same order it was presented from the source.

Review

After material has been written, return to the source to double check for content and properly paraphrased language.

• Information has been gathered from several sources and distilled in your words.
• Quotations are used appropriately and infrequently.
• Words or ideas do not follow the same pattern and order as the source material.

The example above on this page illustrates a common way in which people closely paraphrase content; this one demonstrates how to properly synthesize and paraphrase information.

Detecting

Unlike straightforward copyright violations, close paraphrasing is notoriously difficult to detect; frequently the contributor will add wiki syntax and write in the style of a Wikipedia article. Here are some ways you might detect it:

• Look for disjointed and sudden changes in the tone, vocabulary, and style of content introduced by the same contributor. For example, "The cat is a small predatory carnivorous species of crepuscular mammal. Housecats like to kill mice and bats."
• Look for redundant content; this may be a sign that two or more sources were closely paraphrased. For example, "The cat is a small predatory carnivorous species of crepuscular mammal. Like many pets, domestic cats are carnivorous."
• Look for content that resembles content included in a quotation.
• Examine the talk pages of major contributors and other pages where they have written in their own words, and determine if their article contributions substantially differ in tone, structure, and vocabulary from these discussions.
• Take short phrases from the article and put them in a search engine. Take a look at the results and see if they closely resemble the article.

Addressing

Your approach may vary depending upon the severity of the concern. Here are a couple of ways to manage close paraphrase concerns.

Flag the problem

Insert a close paraphrase template or follow copyright infringement discussion: Your approach here may depend upon the extensiveness of the issues you discover.

• You can use the {{Close paraphrasing}} template, which can be customized to identify the source and to indicate if the source is public domain, to mark it for cleanup:

• However, if you believe that the close paraphrasing in question is so close that it infringes copyright, instead follow the instructions at Template:Copyvio, which may require removing the paraphrasing content until it can be repaired. Unless close paraphrasing is immediately obvious, it is good practice to cite specific passages alongside the corresponding passage from the source on the talk page to highlight their similarity; this will provide objective evidence of close paraphrasing.

Notice to the contributor

It is important to discuss your concerns with the contributor. Many people who paraphrase too closely are not intentionally infringing, but just don't know how to properly paraphrase. It might help to point them to this essay or to the references and resources listed here, which include some pointers for proper paraphrasing.

Example approaches to discussing with other editors

Note: All text in these examples is dedicated by its authors to the public domain

The following example messages can be copied and pasted directly from this page, although you will need to fill in your own example close paraphrases as well as supplying the article's title and the source URL. The messages strive to avoid accusations while at the same time pointing to clear instructions on how to fix errors of this sort. The spaces for examples from the editor's inappropriate text are provided because even experienced or good faith editors may not recognize where the issues lie without them. If there is a passage of several consecutive sentences which is a continuous close paraphrase, this may alone be a sufficient demonstration. Otherwise, showing the pattern in several separated sentences is typically better than offering one, brief example.

Extensive problems

The following example was engineered for cases when the paraphrasing is close enough to require blanking of the article and listing at Wikipedia:Copyright problems. It would not be appropriate for situations where the {{close paraphrasing}} template is used, since rewriting can be done on the spot rather than in a temporary page. You may use this example verbatim, if you wish, but may and should modify it if it is not completely appropriate to the circumstances.

Limited problems

This example was engineered for cases when the paraphrasing is not enough of a concern to require blanking and listing and the {{close paraphrasing}} template is used instead. You may use this example verbatim, if you wish, but may and should modify it if it is not completely appropriate to the circumstances.

A number of Wikipedia policies and guidelines are relevant to this essay. They include:

• WP:ORIGINAL. Policy stating that, while articles should be written in your own words, they should substantially retain the meaning of the source material
• WP:COPYRIGHT. Policy that describes general principles that apply to use of copyrighted work
• WP:COPYVIO. Policy that describes copyright violations and how to address them.
• WP:NFCC. Policy that defines limitations on use of non-free content
• WP:NONFREE. Guideline that expands on WP:NFCC and describes when non-free material may be used under the "fair use" principle
• WP:PLAGIARISM. Guideline that describes the importance of attributing the sources used, even when they may be out of copyright
• WP:MOSQUOTE. Guideline that describes how quotations should be faithfully reproduced, clearly identified as quotations

Several Wikipedia articles discuss related topics such as Copyright law of the United States, fair use, plagiarism, moral rights and paraphrasing of copyrighted material. These may be of interest to editors. However, they may have inaccuracies or omissions, and Wikipedia has a broader aim of providing material that may be used anywhere for any purpose, which imposes further restrictions that are defined in our policies and guidelines.

• Wikipedia:Plagiarism (guideline) — Guideline on Wikipedia's definition of and approach to handling plagiarism to avoid making the work of others look like your own.
• Let's get serious about plagiarism, Signpost (dispatch) — Thorough discussion of Wikipedia's position regarding plagiarism, including how to avoid it, good practices and how to detect it.
• Copy-paste (information page) — Information about plagiarism and copyright concerns with copy-paste, which cannot be done except in the case of short quotations.
• Paraphrasing of copyrighted material (article) — An overview of how close paraphrasing is treated under various jurisdictions' copyright laws.
• Idea-expression divide (article) — Idea-expression divide limits the scope of copyright protection by differentiating an idea from the expression or manifestation of that idea.
• Derivative works (copyright FAQ) — Discussion of derivative work: a copyright concern where something that is "based on and a close copy of" another work.
• Derivative works (talk page) — Discussion of paraphrase and abridgement, a copyright concern.
• Quotation copyright and fair use (essay) — Proper use of quotations in articles, with copyright and "fair use" considerations.
• Substantial similarity (article) — Substantial similarity is the standard developed and used by United States courts to determine whether a defendant has infringed the reproduction right of a copyright.
• How to write a plot summary (essay) — Information about how to write a plot summary for a work of fiction.
• Plot-only description of fictional works — Among other things, discusses how to avoid copyright violations when writing about fictional works.
• Plagiarism detection – includes a list of free and commercial plagiarism detection tools.
• Template:uw-paraphrase - user warning which can be used for users which paraphrase copyrighted material on Wikipedia.

• Paraphrase: Write it in Your Own Words, Purdue University
• How to Paraphrase Without Plagiarizing, Colorado State University