Wikipedia:Wikipedia Signpost/2008-08-11/Dispatches

From Wikipedia, the free encyclopedia
Jump to: navigation, search
The Wikipedia Signpost

Dispatches: Reviewing free images

By Elcobbola, August 11, 2008

Wikipedia's best articles are often enhanced by images. Indeed, the featured article criteria ask for "images and other media where appropriate" and that, as for the use of all images in Wikipedia, they should have "acceptable copyright status. Non-free images or media must satisfy the criteria for inclusion of non-free content and be labeled accordingly." Similarly, the good article criteria require that images be "tagged with their copyright status" and that valid fair use rationales be provided for non-free content.

Images on Wikipedia are classified as either "free" or "non-free":

  • "Free" images are in the public domain or are copyrighted but have no restrictions on derivatives, commercial use and permission for use.
  • "Non-free" images (those with a "fair use" justification) retain restrictions on derivatives, commercial use and permission for use; therefore, they are allowed on Wikipedia only under the restrictive terms of the non-free content criteria.

This dispatch discusses free images, and explains how to ascertain whether or not an image is actually free. A future Dispatch will cover the use of non-free images.

Although all Wikipedia content is expected to have acceptable copyright status, featured article candidates receive particular scrutiny for compliance with the image usage policy. Examining image licenses is not always straightforward. Ultimately, it is a matter of confirming that a copyright tag is present and that the information provided is sufficient to corroborate the tag that has been selected.

Copyright and copyleft

Copyright is a legal protection granting the creator of an original work – for our purposes here, an image – exclusive rights to that work. These rights prevent others from copying, redistributing or modifying the image without the author's permission. Copyright is generated automatically on the creation of such a work.

Copyright holders may choose to relinquish some or all of their rights, for example, by licensing their image so that others may copy, redistribute, or modify it without seeking permission. Such licenses are typically called "copyleft" licenses – a play on the word "copyright". Copyleft images are still under copyright; their creators have merely waived some, but not all of the protection that copyright affords them.

Commonly used licenses include:

  • Creative Commons (CC). These are a range of licenses that govern the copying, redistribution and modification of an image. CC licenses are modular: the "base" license ("CC-by") requires just attribution. This may be supplemented by additional conditions such as SA (share alike: all derivatives must be published with the same license), NC (no commercial usage) and ND (no derivatives). For example, CC-by, CC-by-NC, and CC-by-NC-ND are possible variations. Not all variants are copyleft or are acceptable on Wikipedia (see "common misconceptions" below).
  • GNU Free Documentation License (GFDL). This allows for the copying, redistribution and modification of an image, even for commercial purposes. (Incidentally, this is the license under which Wikipedia prose is published.) Attribution is always required.

Public domain

Works in the public domain are not owned, controlled or otherwise restricted by any person, entity or law in a given jurisdiction. A public domain image may be freely used, altered and published by the public at large without condition.

Generally, an image enters the public domain when it is no longer eligible for copyright protection, usually a certain number of years after its first publication or after its creator's death. The length of time before copyright protection lapses varies greatly from country to country. Because the Wikimedia Foundation servers are located in Florida, images used on the Wikipedia must be in the public domain in the United States.[1] Non-US images hosted on Wikipedia are not required to be public domain in their country of origin provided that they are public domain in the United States. Images hosted on the Wikimedia Commons, by contrast, must be public domain in both the United States and their country of origin; compliance with Commons policy, however, does not figure in the FA or GA criteria.

Copyright terms in the US vary according to several conditions. The most common encountered on Wikipedia are:

  • published works, for which copyright has expired if the image was first published in the US before January 1, 1923; and
  • unpublished works, for which copyright has expired where the creator has been dead at least 70 years.

An image may also be voluntarily released to the public domain by its copyright holder or, in certain cases, may not be eligible for protection in the first place.

Reviewing images

Article reviewers generally need to take into account three aspects:

  • ensuring that policy-mandated elements are present;
  • ensuring the claims of the copyright tag are supported by the source; and
  • consideration of possible legal nuances.

Policy-mandated elements

Wikipedia's image usage policy requires all images to have three pieces of information:

  1. A copyright tag
  2. A verifiable source
  3. An image summary

1. A copyright tag is a template, typically rectangular and appearing towards the bottom of an image page. The tag indicates the image's license or, if public domain, the reason the image is no longer eligible for copyright protection. The {{GFDL}} copyright tag, for example, appears as follows:

This template will categorize into Category:Wikipedia license migration candidates.

2. A verifiable source can be in the form of a simple weblink, citation for the published work from which the image was scanned or the name and method of contact for the author. The format and location of sourcing information on an image description page may vary. Optimally, images will use the {{Information}} template, which provides organized source and summary information. This template is not mandatory, however, and the information may be "hidden" within template boilerplate (example), if present at all.

3. An image summary provides the "necessary details to support the use of the image copyright tag". WP:IUP recommends the following:

  • Description: The image subject
  • Source: The copyright holder of the image or weblink, published work, etc. from which the image came
  • Date: The date the image was created
  • Location: The location in which the image was created
  • Author: The image creator and/or copyright holder
  • Permission: Who or what law or policy gives permission to post on Wikipedia with the selected image copyright tag
  • Other versions of this file: Derivatives of the image, if they exist on Wikipedia


After confirming the presence of the three required elements, reviewers should also examine the source provided. Like prose quotations or statistics, images should have verifiable and reliable sourcing. By their very nature, image copyright tags (especially those claiming public domain) are "material challenged or likely to be challenged" and, consequently, subject to Wikipedia's verifiability policy (WP:V) and the necessity of utilizing reliable sources (WP:RS).

Consider, for example, the following copyright tag:

An image employing this copyright tag would be expected to have a reliable source explicitly indicating the author's date of death or dating the image such that no reasonable scenario would contradict the claim (e.g. the author of a painting dated 1740 could not possibly have been dead less than 100 years).

The following are examples of correctly formatted, verifiable and reliable sourcing:

  • Weblink (example): Images from websites should not link directly to the image itself, but to a page on which the image is used that also contains information to corroborate the copyright tag.
  • Published source (example): Images from published material should contain enough information to identify the specific edition of the book, journal issue, magazine, etc. from which the image came (e.g. merely citing The New Yorker is insufficient).
  • Author contact (example): Images sourced to an author (typically "self made" images) should explicitly indicate the author and provide a link to a user page or other means of contact.

WP:V notes that "the appropriateness of any source always depends on the context". A Geocities site, for example, claiming that an image is public domain will probably not be considered sufficiently reliable to support the claim. Institutional and research sites (e.g. libraries, museums and archival sites such as the Library of Congress) are generally the most reliable.

Legal nuances

Copyright law is often nuanced and esoteric; consequently, there are many concepts of which image authors and uploaders may not be aware. "Derivative works" and "freedom of panorama", two such concepts, can be counter-intuitive and, as such, are a common cause of unintentional copyright violations.

Derivative works
This image of the Fountain of the Great Lakes is a derivative work. The copyright status of the fountain/statue and the photograph itself need to be considered.

A derivative work is a copy, translation or alteration of an existing work – for example, a scan of a page in a book or a picture of a stuffed animal. The Wikimedia Commons' derivative works guideline contains an example situation which explains the dilemma such images pose to Wikipedia:

By taking a picture with a copyrighted cartoon character on a t-shirt as its main subject, for example, the photographer creates a new, copyrighted work (the photograph), but the rights of the cartoon character's creator still affect the resulting photograph. Such a photograph could not be published without the consent of both copyright holders: the photographer and the cartoonist.

Wikipedians or external sources may believe in good faith that a scan, photograph, or screenshot that they have made is an entirely original work, thinking that, because they themselves made the scan or took the photograph, the resulting image is "self-made" and, thus, "free". This is not necessarily the case. Reviewers should consider whether the subject of the image is under copyright – a consideration independent of the copyright status of the image itself.[2]

Although not mandatory, derivative images will, ideally, have summaries identifying the copyright status of both the image and its subject. The image to the right, for example, contains a secondary copyright tag for the fountain/statue. In its case, the image as a whole is "free" and acceptable on Wikipedia, as the subject is demonstrably in the public domain. Alternatively, consider an image of a Batman action figure. Although the image itself could have any copyleft license, the image as a whole would still not be acceptable on Wikipedia, as the figure has not been published with a "free" license.

Freedom of panorama

Freedom of panorama is a copyright law provision that allows for photographs of works (e.g. buildings and sculptures) permanently installed in public places to be freely published, even if the works are still under copyright. Although such an image is still a derivative work (i.e. a translation of an existing work), it does not infringe the rights of the work's author in countries with freedom of panorama. In other countries, however, the derivative image requires consent of the subject's author to be freely licensed.

The United States does not have freedom of panorama, although pictures of buildings are exempt.[3] Hence "self-made" images of publicly-situated works in the United States require consent of the subject's author, as described above. This revision of an image depicting Jaume Piensa's Crown Fountain in Chicago, for example, is incorrectly tagged. As a photograph taken in a country without freedom of panorama (the USA), it would require the permission of the fountain's creator for it to be published with a CC or GFDL license.



Unless an image is deliberately employing pointilism, the appearance of dots when the image is magnified may be a cause for concern.

"Self made" images are generally those which are uploaded by their authors (i.e. Wikipedian-created images). In addition to checking for the policy-mandated elements, it is helpful to consider several aspects pertaining to provenance:

  • Discrepancies: Does the image itself contradict the information available? For example, a car reported to have been photographed in Boston should raise questions if it has a European Union license plate.
  • Metadata: Does the image contain metadata? Modern digital cameras tag images with camera type and other technical data. Images taken from websites rarely contain metadata.
  • Resolution: Does the image have a reasonably high resolution? Modern digital cameras generally produce very high resolutions (for example, a 3.1 megapixel camera produces 2048×1536 pixel images). The more pixels an image has (the higher the resolution), the larger it is. Images used on web pages, therefore, often are low resolution, to reduce load time and to fit within the computer's screen resolution. Consequently, low resolution images (e.g. 300 x 200 pixels) claiming to be self-made should elicit additional scrutiny.
  • Technical quality: Does the subject appear posed, to have been taken in a studio or possess other "professional" traits? Some Wikipedians are indeed professional photographers, but unusually high technical quality should elicit additional scrutiny.
  • Telltales: Does the image have ominous visual cues? Images with watermarks or borders should raise red flags. Additionally, for example, scanned images may contain halftones, small dots which appear as the image is magnified (i.e. zoomed in upon). The presence of half-tones may indicate that the image is not the work of the uploader.
  • User: Is the uploader an established Wikipedian or a "drive by" uploader with few or no other contributions. Always assume good faith, but remember that less established users may be unfamiliar with Wikipedia image policy.

Reviewing images requires common sense. Consideration of provenance is an art, not a science, and the above notes should not necessarily be used as a "checklist". Whereas any one of these considerations may be meaningless by itself, a combination of issues may bring the validity of an image into question. A talk page note to the uploader asking for clarification or a Google images search, for example, may be appropriate or necessary to be more confident that image is indeed "self-made".

Good image
In full compliance with Wikipedia image policy and properly licensed, the good people of Rhinebeck, New York are able to enjoy a sunny day.

The "self-made" image pictured to the right (as of this version) is in full compliance with Wikipedia policy and properly licensed.

  1. Is there a copyright tag? Yes, it asserts that Daniel Case has released the image as GFDL version 1.2.
  2. Is there a verifiable source? Yes, it asserts "self-made"; the uploader matches the author and a link to the author/uploader's profile is included.
  3. Is there an image summary (i.e. the "necessary details to support ... the image copyright tag")? Yes, the image has a complete {{information}} template.
  4. Does the provenance check out?
    • The image is high resolution (1,929 x 1,284 pixels – on the image description page, look below the image itself or in the "Dimensions" field of the file history).
    • The image contains camera metadata (on the image description page, under the metadata header).
    • The image does not appear posed, to have been taken in a studio or possess other such "professional" traits which would raise red flags.
    • The image is dated September 29, 2007 (i.e. well after the claimed license – GFDL version 1.2 – came into existence).
Flawed image
The Japanese Tea Garden is far more lush than the image's summary.

The "self-made" image pictured to the right (as of this version) is in not in compliance with Wikipedia image policy.

  1. Is there a copyright tag? Yes, it is using the {{GFDL-self}} copyright tag (note that, unlike the GFDL example above, this "self" variant begins with "I, the creator of this work").
  2. Is there a verifiable source? No, there is no explicit assertion of authorship, and, accordingly, no means of contacting the author.
  3. Is there an image summary (i.e. the "necessary details to support ... the image copyright tag")? No, the image only includes a description of the image's subject.
    • The image summary is essentially non-existent and, consequently, lacks necessary details. The copyright tag implies the uploader is the "I" in the copyright tag, but explicit indication is needed. Compare with the information present in the example above.
  4. Does the provenance check out?
    • The image is a mid-resolution (at 800 x 600 pixels, it is just under 0.5 megapixels). Although this is a higher resolution than most web images, it is lower than expected and is also a common computer screen resolution (i.e. what one might find at a computer wallpaper archive site).
    • The image does not contain camera metadata.

The verifiable source and image summary elements can, in many "self-made" cases, be reasonably treated as one thing. The uploader (i.e. presumed author) would really only need to add a statement to the effect of "Author: J. Ash Bowie" to the summary to resolve the issue.

Already published

Already published images are those which have been obtained from external websites, published works or are otherwise not the authorship of the uploading Wikipedian. Provenance considerations for these images include:

  • Anachronisms: Particularly relevant to images claiming PD due to age, does the image appear appropriate to its indicated time period? For example, a medieval illuminated manuscript or stained glass window should raise red flags if text therein is in English and not, for example, Latin.
  • Date: Does the date seem reasonable relative to the subject? An image of the Empire State Building (completed in 1931), for example, should not claim a date of 1920.
  • Licensor: Is the source reasonably expected to have licensing rights? Imagine, for example, a website dedicated to World War II era tanks containing vintage images of tanks in action and a general disclaimer that all information presented on the site is licensed as GFDL (i.e. "free"). An image from this site, uploaded to Wikipedia with a full summary and link to the disclaimer seems, prima facie, acceptable. Doesn't it seem implausible, however, that webmasters in the 21st century would have been photographing tanks in battle in the 1940s? Is it not odd that images taken in the 1940s would have a copyright license first released in 2000 (GFDL)? What is likely the case, in this scenario, is that the webmasters scanned the images, and believed the scans were theirs to license. This is not the case, as the scans would be considered "derivative works", as per above.
  • Technical quality: As with "self made" images above, does the image have an expected technical quality? Professional images are not commonly published with free licenses; high technical quality and unusually good vantage points (e.g. when photographing celebrities) may raise red flags. Similarly, images claiming PD due to age are generally of inferior technical quality; "vintage" photographs are typically black and white or sepia, over-exposed and are less sharp than contemporary photographs.
Good image
Hard work and diligence like that exhibited by Mackintosh and Spencer-Smith yields soundly-sourced images.

The image pictured to the right (as of this version) is in full compliance with Wikipedia policy and properly licensed.

  1. Is there a copyright tag? Yes, it uses the {{PD-US}} copyright tag.
  2. Is there a verifiable source? Yes, a full citation of the published source from which the image originated has been provided.
  3. Is there an image summary (i.e. the "necessary details to support ... the image copyright tag")? Yes, the citation contains the publication date (1920), which supports the copyright tag's assertion of first publication before January 1, 1923.
  4. Does the provenance check out?
    • The image has expected technical qualities. The image is black and white and generally appears to be old.
    • The image has reasonable subject matter. Mackintosh and Spencer-Smith were indeed in Antarctica before 1920 (the reported publication date).
Flawed image
Lacking a verifiable source and image summary, Emperor Valerian is humiliated in the ensuing chaos.

The image on the right (as of this version) is in not in compliance with Wikipedia image policy.

  1. Is there a copyright tag? Yes, it is using the {{PD-art}} copyright tag (claiming the image is in the public domain because the author has been dead more than 70 years).
  2. Is there a verifiable source? No, a source (e.g. web link or published source) has not been provided.
  3. Is there an image summary (i.e. the "necessary details to support ... the image copyright tag")? No, the information provided is not adequately supported. The names of the uploader and asserted author do not match, which indicates the image is not self-made and, thus, there exists an external (non-Wikipedia) source that needs to be cited.
    • Without a source, we cannot confirm that the asserted author (Hans Holbein the Younger) is indeed the original author.

Although this is likely public domain, verifiability, not truth, is the threshold for inclusion. Without a source confirming the author, this image could just as easily be a contemporary work.

Common misconceptions

  • Copyright subsistence: To reiterate a concept mentioned above, images with copyleft licenses are still under copyright. Retention of any right (e.g. the requirement of attribution) retains the copyright. An image is either copyrighted or in the public domain. There is no middle ground.
  • Government hosting: Works of the federal government of the United States are not generally eligible for copyright. It is a common misconception, therefore, that an image on a federal website, or in a federal report, etc. is public domain. Such images should not be assumed to be public domain in the absence of explicit assertion of federal authorship or a general disclaimer that all images on the site or in the report, etc. are public domain. The FTC complaint against Movieland, for example, contains screenshots of copyrighted software. Although the FTC is a federal agency and the text of the report is public domain, the screenshots are derivative works to which the federal government does not have rights.
  • Imprecise disclaimers: It is not uncommon for websites to contain disclaimers to the effect that "material herein is free for all to use", "images may be freely published", etc. Wikipedia uploaders, if unfamiliar with licensing or copyright, may select a GFDL, CC or {{Copyrighted free use}} copyright tag. These disclaimers are not acceptable, as the source has not fully articulated what "free" means (e.g., whether derivatives may be created).
  • All licenses are not created equal: Copyright license variants and versions matter. Both CC and GFDL have version numbers (e.g. CC-by 2.0 or CC-by 3.0 and GFDL 1.1 or GFDL 1.2). Although the version number does not impact the image's acceptability on Wikipedia, they are indeed legally different and care should be taken to ensure that the correct version is reflected by the copyright tag. CC licenses, additionally, are particularly troublesome, as not all variants are "free". Wikipedia uploaders, if unfamiliar with CC nomenclature, sometimes upload, for example, an image licensed CC-by-NC 2.0 as CC-by-SA 2.0 because it seems "close enough"; it isn't. Per WP:IUP, "Images which are listed as for non-commercial use only, by permission, or which restrict derivatives are unsuitable for Wikipedia and will be deleted on sight."
  • Creation is not publication: Sources often only provide the date a work was created. Creation is quite different from publication, which is "the distribution of copies or phonorecords of a work to the public by sale or other transfer of ownership, or by rental, lease, or lending".[4] Taking a photo and placing it in an album or hanging it on a wall, for example, is not publication. Producing a painting or sculpture is not publication. Using the photo or an image of the painting or sculpture in a publicly-distributed book, newspaper, journal, postcard, or other such medium is publication. For PD claims based on the date of first publication, the source needs to indicate the actual publication date.


  1. ^ WP:IUP: "U.S. law governs whether a Wikipedia image is in the public domain"
  2. ^ Copyright law contains a provision and exception for "article[s] having an intrinsic utilitarian function". A picture of a car or a chair, for example, would not be problematic. See the Commons guideline for elaboration.
  3. ^ 17 USC 120(a)
  4. ^ United States Copyright Office (2006). Copyright Office Basics. Retrieved August 1, 2008.

See also

Also this week:

Growth study — Board Nominating Committee — Greenspun project — WikiWorld — News and notes — Dispatches — Features and admins — Technology report — Arbitration report

(← Previous Dispatches) |width="20%" align= "center" valign="top" |Signpost archives ||width="40%" align="right" valign="top" | (Next Dispatches→)