2010 Wikimedia Study of Controversial Content: Part Three

The following is part three of the Draft Study on Wikimedia Controversial Content by Robert Harris and Dory Carr-Harris. Comments on the draft or questions are highly encouraged, and should be placed on the talk page, but for now please refrain from revising the draft directly. The preliminary statements to the recommendations on this page are located at 2010 Wikimedia Study of Controversial Content and Part Two, and they should be read for full context before commenting.


We want to thank all of you who have contributed to this discussion so far. It has been very instructive for me to absorb (even if it occasionally did seem as though I was watching Question Period (a British parliamentary tradition) in my native Canada every so often (“Some Honourable Members: Oh, oh”, for those of you addicted to Hansard).

I am not trying to cut off conversation on the talk page associated with Part 2 of the study, nor on the on-going conversation on the Talk page attached to Part 1 of this study, but here is part 3 of the study. Part 3, as you will see, contains no recommendations – merely some observations of topics that were too briefly covered in our discussions so far, in our opinion. They include a slightly fuller discussion of the relationship between Wikimedia projects, children and their families; an attempt at an explanation of why we think images sacred to one group or another should be treated differently within Wikimedia projects than other kinds of images; and a noting of what we consider to be something worth mentioning – the difference between the two mandates Commons has taken on for itself – of providing images to support the other projects, and to make an image bank for the world – and the significance of those different mandates.

We are presenting this study to the Board of the Wikimedia Foundation, more or less as presented here, on October 8th. I say more or less, because it is possible we might make some small changes in the manner and order in which we actually present the material. But the recommendations we have made will not change – what you have read is what they are. We, of course, cannot say what will happen to those recommendations after we present the Report to the Board, only to remind you what the board itself said when it introduced the Study, contained in its Qs and As:

What will happen after the board receives the recommendations?

The board will discuss the recommendations at its fall meeting. Then it will talk with the community. Nothing will happen without lots of discussion.

The issues raised in this study are not easy ones to either grapple with or resolve. They revolve around issues that are fundamental and content which often is emotional. We thank you for the care and diligence with which you have participated in the discussion, which I’m sure is far from over, but we would like to think, a little closer to potential resolution.


The question of the relationship of the Wikimedia projects to children and their parents is a complex one, with many facets, that has come up periodically in the history of the projects (most recently, this summer, as part of a discussion about Simple language versions of the Wikipedias). In some ways, children seem to be just another community that Wikimedia has considered serving, but, to our mind, there are at least three aspects of the relationship of the projects to children that make it a unique potential venture.

The first is the definition of child. Depending on the jurisdiction, a legal minor can be anyone from a newborn to a 17-year old (under the age of 18). For educational purposes, that would mean that a “children’s” Wikipedia (if we equate legal minor with child), would have to accommodate the needs of a pre-schooler up to someone in first-year university, if we were just to have one version of the project. So, obviously, serving the needs of “children” could not be accommodated by a single version of the projects. If we were to decide to refine that definition, and concentrate on dealing with pre-pubescent children (since in many cultures, the cultural definition of adulthood is the onset of puberty), we would first have to deal with the fact that the onset of puberty is occurring at a younger and younger intellectual age in many parts of the contemporary world,[1] and grapple as well with the questions of appropriate content for this age group. All of which is to say, that, unlike other potential projects, the ambit for a project for “children,” as a basic first step, would need to be defined carefully before such a project could move forward.

As well, we assume that a children’s encyclopedia might not be able to adhere to the Wikimedia principle that projects should be created by their users (although this may be simple ageism; we are constantly surprised by the youth of some of our most outstanding contributors to the regular Wikipedias). However, if we do accept that a child’s encyclopedia might have to be written by adults, it raises questions unique to Wikimedia projects that would have to be debated and settled.

And finally, children are not quite fully-autonomous citizens in legal terms, and where their rights and the rights of their parents or legal guardians coincide, diverge, are in harmony, or are not, is an extremely complex moral and legal question which we note we have not yet studied at any level of detail. We have accepted the “convention,” if you will, (which has legal force as well) that parents enjoy some rights over their child’s behaviour, and will, and do, exercise those rights, in varying degrees among different cultures and societies throughout the world.

The major recommendation we have made to deal with children and their parents is our recommendation to allow users (at their discretion, and only for their personal use) to place some images (of sexuality and violence) in collapsible galleries so that children (for example) might not come across these images unintentionally or unexpectedly. As we noted in our section on basic principles, we did so because we believed it would show some basic respect and service to one group of our users (those worried about exposure to these images) without compromising the different needs and desires of another (those desiring, even insisting, the projects stay open).

We have chosen to nominate images of sexuality and violence for this treatment because it is our observation that concern about these types of images (and their exposure to children) is widespread in different cultures around the world. As we noted in one of our responses to comments on Part 2 of this study, it is often implied (or stated) that this type of restriction is a “biased” attempt by repressive forces in North America, illegitimately uncomfortable with expressions of sexuality, to force their views on the rest of the world. In truth, contemporary North American society is one of the most permissive in the world, compared to most other cultures – in China, the Indian sub-continent, or Muslim countries – and if anything, the rest of the world sees exactly the opposite when they view the major North-American dominated sites of the Internet – a “biased’ attempt to force extremely permissive views of sexuality on the rest of the world. We feel our recommendation around images of sexuality and violence strikes a balance and compromise between these two points of view, by focusing on the principle of individual choice as the determining factor in the use of the tools we are suggesting we offer our audiences.

However, it is our belief that our relationship to children is most powerfully presented in positive, not negative, terms, ie in determining what children should see, not what they should not see. We note that in many North American communities, traditionally, the public library was seen as one of the “safest” places for a child, in terms of their exposure to wanted and unwanted material. However, as we learned from our conversations with librarians, the library ideal (if not always the policy in every individual library) is to restrict no material to children – the ideal is that there is one library, equally open to all users identically. The squaring of this particular intellectual circle (that libraries are considered safe for kids, even though they are relatively unrestricted in their access to their collections) is, we believe, because of the many positive programs that public libraries offer to children – reading groups, kids’s collections, specially-designed computer programs. We believe that this is a model Wikimedia might choose to emulate. If the community wanted to make serving children a priority, policies on what content is to be restricted for children is not enough. An active creation of projects designed for children would be, in our view, a more effective step. However, we note that there is much to be decided upon if such a project was to be initiated that we have not fully studied – whether the Simple Wikipedias could serve as a starting point for these projects, whether partnership with other groups might be the best way to proceed, etc.

We will also note in this regard that unlike many Western-inspired societies where access to kid-styled information is not lacking, in some cultures and societies, this is far from the case. In India, for example, basic sex education is not provided for in the school system, a topic of concern and contention within Indian society. A well-presented Wikimedia project for children that dealt, appropriately, with topics of sexual education, might have a profound, and positive, value for such societies. Even for Western societies, we have had expressed to us several times that rather than seeing the issues of children, Wikimedia, and sexual topics always expressed as a negative in the community (i.e. in discussions of how to prevent children from inappropriate contact with this material), a well-presented Wikimedia children’s, or teenager’s guide to sexuality (based on Wikimedia principles of information access, verifiability, etc) might be a significant improvement on other forms of online education currently available.

Images of the “sacred”

We have suggested a regime around management of “controversial” sacred images different for that of other kinds of controversial images because, in our mind, these kinds of images are different in kind from other controversial images. Despite surface similarities, sacred images represent a different intellectual process of designation, and demand a different form of resolution, in our view.

The reason we think so is that we believe conversations and decisions about images considered sacred by one group or another are always conversations about two different scales of values held by different groups of people, where the key to resolution is the attempt of one group of people to understand and respect the values of another. We believe this is different than the case for other forms of controversial content.

For example, if I show you a video of a beheading on the Internet, it would not surprise you to find that there is a variety of opinion about its appropriateness for use on Wikimedia projects. The notion that this is a potentially controversial image is obvious and clear without explanation. We may fall at varying points on the scale of how controversial it is, and how severely its use should be limited (if at all) on Wikimedia projects, but they are points on the same scale. If, on the other hand, I show you a picture of a seated, bearded Persian man, wearing a Fez, taken in the late nineteenth century, the image seems neutral in and of itself. Nothing in it is inherently controversial. However, when I tell you that it is an image of Baha’u’llah, the founder of Baha’i, and that its display is considered inappropriate by members of the Baha’i faith in certain circumstances, you, in effect, are being confronted with a different definition and scale of values used to evaluate the image. This is true, whether the image in question is of the founder of Baha’i, or of Mormon Temple Garments (considered sacred by members of that faith), or images of Muhammad. With images of this sort, rather than confronting a series of responses arrayed on the same scale of values, different groups of people judge the image using different scales of values, and the key to reconciling them is for each group to absorb and understand the values of the other, after which, and only after which, a determination of action around these images may be taken.

Note that this situation does not make it any easier to decide questions around the use or restrictions on usage for these kinds of images, (in fact, we would argue that it makes them more difficult.); however our basic principles, we believe, still apply to these kinds of images. We start with a bias towards openness but agree to limit that openness, based on respect for our users, as little as possible. That is why, we believe, potential decisions on the restrictions of these types of images must be decided by individual users and why we have recommended that registration be necessary to affect these images. A more general prohibition of them, given their nature, would seem to be moving too far, in our opinion, in the direction of general restriction of the projects.

The reason we have come to that conclusion stems from our observation that pluralistic, multi-faith, and secular communities are a common feature of many societies around the world today (and the values of the international virtual community of the internet) and that in these societies, questions of the appropriateness of the display of “sacred” images, as defined by one community, are inevitably decided within the context of other communities who do not value the same images in the same way. It is about the struggle between the rights of some individuals to define the limits of appropriateness of some images for themselves and others, versus the rights of others to know, and the question of the amount and quality of respect that should be offered to each by each other.

Commons Dual Mission

We noted in passing in Part 2 that Commons has a double mission: to act as a resource from which other Wikimedia projects may find suitable images, as well as to be a freely-licensed source of educational images for anyone seeking such images, for the world, in effect. These two purposes have been part of Commons raison d’etre since its inception, and have existed hand in hand ever since. However, it is our belief that these two goals are increasingly at odds with each other, contradictory to each other in some fundamental ways, and that as Commons matures and goes forward, these contradictions will become more apparent, and may potentially be a liability for Commons’ increased and future health.

We have no experts to cite as references for this opinion. It is merely an opinion, which is why no formal recommendations have been made on this subject. However, we feel we would be remiss not to share our views on the subject, and our rationale for arriving at them, as part of this study.

First, as regards Commons’s growth. The statistics we cited in Part 2 showed that Commons is accessing new images at the rate of approximately 7,000 per day at the moment (it varies from month to month, sometimes higher, sometimes lower). At that rate, Commons will access over two and a half million images every succeeding calendar year. That means it will double its current eight million image bank (which took it six years to amass) in a little over three years. As well, we note that procedures to improve the accession of video on the site are being developed, and that the wide presence of video on the projects will present new challenges when it comes to certain forms of controversial content, if not many other forms of content as well. (We note that YouTube, has the strictest rules on accession of content of all the major, big-ticket, widely-used sites: no explicit sexual nudity of any kind is permitted on YouTube; Flickr deals with the extra emotional content of video by restricting all videos on their site to 90 seconds.)

We note that Commons’s growth these days is far more rapid than the growth of the major Wikipedias. And although we have no current numbers to back this up, we assume that that means that the number of images on Commons used by sister projects, as a percentage of the total number of Commons images, will inevitably drop as these trends continue. What this means is that Commons, we believe, will begin, more and more to be seen primarily in its second role, as an image bank for the world.

Why we feel this might conceivably be a problem for the community, is our belief that educational scope will be increasingly difficult to appropriately define as the image bank becomes focused on its wider mission of collecting all potentially available freely-licensed and public domain images for general distribution. If the goal of Commons is to collect all freely-licensed images, with the implication that an image is “educational” by its very nature of representing some facet of human experience, any other more tightly-restricted definition of educational, will inevitably be eroded on the project. In this regard, we believe that the two missions of Commons are actually contradictory in one key regard. Where the mission to provide images for the other projects is centrifugal, centered in on the selection of a number of images for an immediately identifiable purpose, the mission to become an image bank for the world is centripetal – focused toward the collection of a wide mass of images, where limits are unnecessary, perhaps even counter-productive, and totality is a goal. The concept of educational scope, we think, becomes increasingly difficult to define as these archival instincts take over Commons, and creates the possibility that the project might take on characteristics different from the ones originally intended for it.

Even if there is agreement that such a universal, “big-bucket” image bank is desirable, where the collection of every image of everything becomes a goal, there is a second question of whether the Wikimedia Foundation is the appropriate institution on the Internet to host such an archive. If Wikimedia’s goal is to be educational, thus, in some way selective of information, perhaps this is a project that fits more appropriately outside of the Foundation’s ambit.

We realize that these are neither simple, nor uncontroversial questions, and please note that we are making no recommendations at all one way or another about these issues, but we are raising them, as we felt that it would be remiss for us not to do so, and we wanted to bring them to the Board’s and the community’s attention as a tangential sidebar to our more immediate mandate to deal with controversial content on the projects.

Relationship with our Audience

We note with regard to the recommendations contained in this study that a new Chief Community Officer has recently been hired within the WMF, and we encourage the community to support that department, especially in its work to create relationships with readers, in light of some of the principles and recommendations we have made here. Whatever is eventually decided about these recommendations, we believe that Wikimedia projects in general should always strive to communicate their aims and intent in as friendly, open and communicative a manner to our audiences as possible. Although every new project within the Foundation or the community involves time, and resources, the notion that Wikimedia should have a blog where audiences can communicate directly with the projects seems like a good one to us, as well as, perhaps, a re-evaluation of the language used in our disclaimers, which we believe might be able to be re-worded in a slightly different tone. Obviously, as one of the posters on Part 2 noted, the bottom line is that Wikimedia projects are serving our audiences exceptionally well, and the proof is the usage statistics that put Wikimedia projects among the internet’s most popular sites. That is our basic starting point, but there is no reason not to follow it up with a regime of connection with our audiences that further cements their appreciation of what we – you – do.


  1. Some Girls' Puberty Age Still Falling, Los Angeles Times, August 9 2010