One of the Nine Theses on Wikipedia series

☞ Read this essay on Wikipedia.
☞ Comment on Wikipedia. (And elsewhere on Wikipedia.)
☞ You can also leave comments below.

2. Enable competing articles.

Neutrality is impossible to practice if editors refuse to compromise—and Wikipedia is now led by such uncompromising editors. As a result, a favored perspective has emerged: the narrow perspective of the Western ruling class, one that is “globalist,” academic, secular, and progressive (GASP). In fact, Wikipedia admits to a systemic bias, and other common views are marginalized, misrepresented, or excluded entirely. The problem is that genuine neutrality is impossible when one perspective enjoys such a monopoly on editorial legitimacy. I propose a natural solution: Wikipedia should permit multiple, competing articles written within explicitly declared frameworks, each aiming at neutrality within its own framework. That is how Wikipedia can become a genuinely open, global project.

Note: The first four theses all concern different aspects of neutrality. This involves some repetition and expansion of analysis, because the issues involved are so central and important.

The Problem

Wikipedia was started by two libertarians devoted to openness and freedom. For us, it was a foregone conclusion that it would be pluralistic. Of course we wanted Wikipedia to represent a wide variety of views—the more, the merrier. We wanted the whole world to come together and articulate their opinions with the best sources, allowing others, holding quite different views, to do the same. We originally expected there to be a global smorgasbord of thinking recorded, reflecting widely divergent politics, nationality, religion, and more. Wikipedia was supposed to be like a big ethnic food fair. It’s food (for thought) from everywhere in the world. It doesn’t matter who you are—you’re guaranteed to be puzzled, surprised, and delighted. Your taste buds will be tantalized, and you will inevitably enjoy yourself (barring gastrointestinal complaints). This ideal is reflected in the Wikipedia logo.

But that is not how it works today.

To some extent, Wikipedia even admits this. For many years, Wikipedians have wrung their hands over their own “systemic bias.” [a] They are—and this is by their own account—too white, male, technically inclined, formally educated, English-speaking, younger, etc. Also, apparently, they have too many Christians.² On their own telling, it is a terrible thing that women and people from the Global South are underrepresented in Wikipedia’s ranks. But they are more right than they know.

What the authors of the “Systemic bias” [a] page seem to overlook is the fact that articles take little or no cognizance of anyone’s concepts, doctrines, theories, and so forth, except as represented by a very narrow slice of Westerners. I would describe this thin slice as globalist, academic, secular, and progressive (GASP). This is the true systemic bias of the platform.

“Perhaps,” the GASP advocates might respond,

but what’s wrong with that? Who better to represent the broad assortment of views on our diverse planet? ‘Global’ means ‘not provincial’. We may be tolerant globe-trotters, but it is good not to be a bigoted rube. Academia stands for objectivity and rigor, and we have that in spades. Secularism is not biased in favor of any one religion; we carefully study and document them all. This is a good thing. As to progressivism, reality is biased in favor of progressive ideas; progressives are not biased by any outmoded old ideas. One wants an encyclopedia to be progressive.

This is precisely how many Wikipedians think. In fact, however, Wikipedia is subject to a syndrome of related biases, represented by the handy acronym GASP. Let us take each letter in turn:

• Globalism, as ordinarily understood, is the view of a remarkably provincial group: typically wealthy, university-educated, and concentrated in a few cities in Western Europe, the coastal United States, and the Anglosphere. Most people on the globe are not globalists.
• Academia, for all its virtues, reflects peculiar assumptions not shared elsewhere. In many fields, especially the humanities, there is a dominant philosophical outlook: secular, progressive, relativistic, and now often hostile to most Western traditions. Objectivity has increasingly given way to activism; the careful rigor that once defined scholarship has eroded.
• Secularism itself is hardly neutral: most people, including many of the most intelligent, are religious. This is a worldview alien to most of humanity, one that scorns all faiths or feigns a perfunctory respect in order to treat them clinically, taking their claims seriously only as objects of study.
• Progressivism, too, is not some inevitable, universal norm or default position. It is at bottom a parochial ideology of Western “elites,” drilled into students at a small class of expensive institutions.

We will elaborate this analysis more in Thesis 4, but let us begin by discussing one of these four. What might be biased about “academia”? According to one interesting Wikipedia “essay,” [a]

If a scholarly claim is principally unworthy of being taught at Cambridge, Harvard, Oxford, Princeton, the Sorbonne, and/or Yale [CHOPSY], then it amounts to sub-standard scholarship and should be never considered a reliable source for establishing facts for Wikipedia.

This is called “the CHOPSY test.” Although not official, this is the common and ruling view, often cited in support of edits. Basically, if a claim—or the source of that claim—would not pass muster at one of the elite Western universities, then it is considered “not a reliable source.”

The proper evaluation of such a policy is obvious to any fair-minded, educated person: This represents a very definite kind of bias. If you are tossing out sources because they are not used at Harvard, then you are going to have a deeply elitist bias. Having spent around twenty years of my life studying and then teaching at institutions of higher education, I can easily anticipate the scorn with which this will be met (reflected by the “Academic bias” essay [a]):

Well, you may not like “GASP,” but what do you propose that we put in its place, for purposes of editing a serious, intellectually respectable encyclopedia? We are the professional intellectuals and writers, and yes, we reflect the views of the leadership of a world that is, indeed, increasingly global and driven by progress. We are open to a wide variety of views, but we do not check our brains at the door, and we do not suffer fools gladly. Knowledge is our business. If you want rational analysis of the facts, then ask our experts.

There is, however, a sensible—even obvious—response to this. Wikipedians claim to be tolerant of a wide variety of global views. Why not actual representatives of those views, citing the sorts of sources they wish to cite? Those who dwell outside of Western Establishment bastions are not idiots just because they do not mouth the pieties of GASP. Some of them can write very well. There are other traditions, you know. They could write for Wikipedia, if you let them. But such true openness and genuine tolerance is unacceptable, precisely because those other traditions fail to pay exclusive homage to GASP sources, through which—Wikipedians imagine—all the benefits of global civilization flow.

A couple of examples should make it clearer how this attitude works out in practice.

Wikipedia has an article titled “Yahweh.” [a] Now, as I am a Christian,³ “Yahweh” is the name of my God. My observant Jewish friends would say the same (though they would not utter the word itself, which is called “The Name,” or in the Hebrew transliteration, Hashem, since it is sacred to them). The repeated uses of the phrase “the LORD” in the Bible are translations of the name of God.

But in the Wikipedia article, [a] we read that Yahweh⁴

was an ancient Semitic deity of weather and war in the ancient Levant, the national god of the kingdoms of Judah and Israel, and the head of the pantheon of the polytheistic Israelite religion. Although there is no clear consensus regarding the geographical origins of the deity, scholars generally hold that Yahweh was associated with Seir, Edom, Paran, and Teman, and later with Canaan. The worship of the deity reaches back to at least the early Iron Age, and likely to the late Bronze Age, if not somewhat earlier.

According to Wikipedia, Yahweh was (past tense) one god (lower case) in a whole pantheon, the chief god in a polytheistic religion. The article thus presents as uncontroversial fact a theory that is held by Bible critics. The claim that Yahweh was a tribal war god is not a neutral, historical fact, but a modern theory, rejected by many of the most deeply erudite Bible scholars around the world, Jewish, Christian, and Muslim.⁵ But to Wikipedia, the claim is treated as “neutral.” The page’s chief maintainers do not tolerate [a] internal debate on the matter. But the article’s stance certainly is not neutral, precisely because it deliberately ignores the majority view on the topic named by the title, a view taken by the billions worldwide who worship Yahweh.⁶ Even the views of serious scholars critical of the supposed secular “consensus” are omitted and treated with scorn.

On the talk page, one gatekeeper writes, “This article is neither about Judaism, nor Christianity. It is an article about Ancient history.” And later, “The Bible isn’t a valid source for evidence of authenticating history. See WP:RSPSCRIPTURE. [a] … There is no Biblical perspective upon Yahwism [i.e., the religion of the ancient Israelites of the First Temple period].” This is convenient for those who like the article in its present state; it means Wikipedians who want to add the Jewish or Christian perspectives about Yahweh are simply not welcome to work on this article, despite the fact that it is indeed the name of their God. They are instructed to proceed to articles titled “God in Judaism” [a] and “God in Christianity.” [a] In the latter, the name “Yahweh” does not appear until some 1,400 words into the article. Hence, the view about the topic described as “Yahweh,” according to the largest religious grouping of people whose God is Yahweh—the Christians—is systematically marginalized by a comparatively tiny minority of gatekeepers.

Here is a different sort of example of Wikipedia’s cultural bias: “Chennai.” [a] An Indian acquaintance of mine told me that, from an Indian point of view, while it’s a good article in many respects, it is strange. It dwells on topics and attractions perhaps of interest to British colonials and Western tourists, but it does not provide all the sorts of facts and analysis that a native would want. Also, the current lede and history sections are “colonial-centric.”

They treat the British East India Company’s arrival like an origin story, as if nothing of importance existed at the site of Chennai before that. Yet, long before the British showed up, the area was already part of major South Indian empires: Chola, Pandya, Pallava, and later, Vijayanagara. There were established settlements, temples, trade routes, and a distinct Tamil identity in the area. The current lede reads as though Chennai began with colonialism. The problem in this case, it seems to me, is not factuality but selection or emphasis. Perhaps it is true that the history of the modern, Westernized city began with the British East India Company. But from an Indian point of view, many centuries of history are glossed over.

The point is that different groups of people find different facts relevant to emphasize, even in cases when all such facts are, in some sense, “neutral.” An article written exclusively by Indians, using Tamil sources, would probably look rather different and be of more use to Indians—even if it were scrupulously neutral.

Skeptics might protest: Where are these robust alternative traditions capable of supporting excellent encyclopedia articles? Any such argument, however, would evince appalling ignorance of the mere existence of independent intellectual traditions in many places in the world. China and India have truly ancient intellectual traditions and quite active practices of journalism and education; while these are influenced by Western practices, they are not merely appendages. The same may be said for Japan, Iran, and Arabic-speaking centers of culture. For the rest, there are universities and journalism all around the world. I would say there is a vast untapped demand of interest in knowing what the world looks like from their point of view—unfiltered, unpatronized, and unbowed by the opinions of snooty do-gooding Westerners. I dare say that, if given an opportunity to speak for themselves in English, we would find many capable scholars (professors and students) as well as journalists in those countries interested in developing neutral encyclopedic content that present the world anew from within their unique frameworks. If, somehow, they did not have to worry about whether their habits, scholarship, and reporting passed muster with the GASP crowd, they would be far more motivated to get on board. What they produce would doubtless be deeply fascinating.

In short, then, Wikipedia is written by a particular kind of person. Only by pretending that his perspective (i.e., GASP) is normative—vanilla, factual, or neutral—can such a person claim to be the “voice of the consensus” (see Thesis 1), the best judge of “reliable sources” (see Thesis 3), and the arbiter of which views are really “neutral” (see Thesis 4).

To put it more briefly: At present, Wikipedians offer their own well-managed, curated perspective on global opinion. But any opinion outside of that perspective is an unwelcome “minority or fringe view”—as global opinion usually is. They will be silenced, should they have the audacity to speak for themselves. Wikipedians will speak for them, or not, thank you very much.

Wikipedians should face up to these hard facts. It is time to record the glorious chorus of worldwide voices. I call on Wikipedia to become the global project it was meant to be, supporting the views of all of humanity—not just those of a narrow, snobbish Western “elite.”

The Reasonable Solution

But if we wish to record the chorus of worldwide voices, then—how?

In the spirit of one of my old rules (“Be bold” [a]), I submit that Wikipedia should become more open:

Permit multiple, competing articles per topic.

As the need arises, people who find the current article to be biased, factually incorrect, badly organized, etc., would be able to start competing articles if they wish, on an ad hoc basis. At the same time, those who wish to write articles specifically for school children, or specifically for experts, etc., would be able to do so. Such articles would be added to the main namespace as alternatives when they were rated highly enough. To develop this solution, let me share some history and then talk about implementation.

This proposal of multiple, competing articles per topic was discussed when we were planning the project that became Wikipedia. One of Jimmy Wales’ partners at Bomis, the parent company of Wikipedia, was Tim Shell. Early on, Tim championed the idea of multiple articles for Nupedia, the predecessor to Wikipedia. I disagreed (I was editor-in-chief), and Jimmy backed me up; thus Nupedia and, later, Wikipedia became one-article-per-topic. Our reason was that there were not enough people involved to write many competing articles; if volunteers were competing to write articles on popular topics, they might not spend enough time on articles about the long tail of less important topics.

Now, did you catch that? Wikipedia allows only one article per topic mainly because, 25 years ago, there weren’t enough writers. Obviously, that problem is long gone. Wikipedia is now the biggest reference site in the history of the world. There is no shortage of people willing to write for a sane and truly open Wikipedia. The project would be absolutely flooded with new writers if the dominant, established editors were not so difficult and did not constantly chase away and block the untrained and undesirable newbies (see Thesis 8).

Wikipedia could recruit tens or hundreds of times as many writers as it has now. All the project would have to do is to allow a diverse humanity to write diverse articles within diverse frameworks.

But, you ask, how could that possibly work? I think there are many possibilities, but, having given this some thought, I propose seven organizing principles.

(1) New articles begin life in the Draft: namespace, or, possibly, a new namespace. So, effectively, they are hidden from the public, to start, and no one need be alarmed about the inclusion of total nonsense on the website.

(2) Articles automatically move to the main namespace when they meet certain objective criteria. Wikipedians would have to hash out exactly how this would work. Here are a few ideas that might be part of the mix. Each is worth debating. (a) The article is different from existing articles by at least 25%. (b) The article should be at least one-third the length of the main namespace article (if there is any), or 2,000 words, whichever is shorter. (c) The article must receive a minimum rating score across a diverse body of human raters (see Thesis 6); or, alternatively, the article must be rated “Approved” by an agreed-upon (i.e., suitably neutral) AI rating system. (d) The article should have at least two sources. (e) There should be at least three contributors (who have made reasonably substantive edits on the articles). I personally am not wedded to any of (a)-(e). These are just ideas.

(3) The article creator determines who works on the article. Without providing any explanation, the article’s first author would decide whether the article’s roster of authors is (a) “by approval only,” i.e., only explicitly included accounts can contribute; (b) “filtered,” i.e., anyone may contribute, except those explicitly excluded from authorship; or (c) “open.” The article creator would also be ultimately responsible for maintaining the lists mentioned in (a) and (b). If an article is marked as open, then it cannot later be changed to by approval only or filtered; but a more restricted article could be made open.

(4) Competing articles about the same topic are distinguished by differing frameworks. The article originator must explicitly declare a framework, a summary form of which will be placed at the top of each article that follows it. This is a particular combination of (a) audience, together with broad national and/or other intellectual tradition(s); (b) acceptable and unacceptable sources; and (c) “Overton window,” i.e., which opinions, broadly speaking, are viewed as being “within the range of respectability.” While all articles will be expected to be neutral within their framework, they will inevitably differ in what range of views are treated (see Thesis 3). Anyone may invent a new framework, and there is no need to have just one article per framework or to create a bunch of editorial committees.

Example frameworks are given in the following table:

Frameworks should not be considered to be distinct encyclopedias or restricted editorial groups. There can be multiple articles on the same topic and name started in similar frameworks.

Frameworks are not intended merely to codify biases. The goal is to acknowledge perspective frankly—a thing that Wikipedia at present both admits and refuses to admit, as we have seen—while maintaining neutrality within clearly defined boundaries. Articles will continue to be subject to a reinvigorated neutrality policy; see Thesis 4. For those who still care about strict neutrality, one of the frameworks I would encourage using is the “Strict neutrality framework.”

For example, one can easily imagine articles written about “Global warming.” The current article (on “Climate change” [a]) works within the “Status quo framework,” asserting in Wikipedia’s own voice that anthropogenic global warming is an uncontroversial fact and that “contrarian” and “denier” voices are merely part of a manufactured controversy. This article, as it stands, is not neutral, and it needs work. Another article, written within the “Strict neutrality framework,” would eschew any particular view on whether there has been global warming and what its cause might be; it would cover with roughly equal attention the views of “climate change activists” and “global warming skeptics.” No view would be asserted as correct in Wikipedia’s own voice. Or suppose there were frameworks devoted to the Democratic and Republican Parties, respectively; while there would be strong emphases on climate change activism, in the former case, and skepticism, in the latter, both would be expected to range more broadly and to avoid taking definite positions, in order to remain neutral.

As you can see from the example, the notion that articles are always found within a “framework” does represent a concession to realism: it means Wikipedia would officially concede the obvious, namely, that different segments of a widely divergent humanity do take different approaches to neutrality, even when they are doing their best to follow the policy; again, the Thesis 4 discussion goes into detail on this point. It also means that we admit that people of widely divergent viewpoints cannot reach a consensus, even when they are sincerely aiming at neutrality.

Let us be clear, however. Under this system, neutrality would still require that all disputed views are to be attributed to their representatives. Thus, a “Reformed and evangelical” article titled “Salvation” [a] should not simply state what salvation requires, in its own voice, but according to a certain Bible writer, theologian, or Church council; and there must still be space made for alternative views. The amount of space apportioned to other views might differ based on the “Overton window,” i.e., range of discourse (a concept more clarified in Thesis 4). An article about firearms, describing itself as “French” (although written in English), might accord little space to the traditional American view on gun rights—because that view is not widely held in France.

(5) The Arbitration Committee and other bureaucratic groups must respect the rules of each framework, although, presumably, they must also apply some more general rules. In other words, in the same way that ArbCom [a] currently respects (and enforces) rules described by the “Status quo framework” above, so also it would respect rules described by other frameworks. Obviously, there would have to be a new kind of rule, i.e., rules concerning frameworks in general. One example rule might be: Multiple articles may be started with very similar frameworks. This might be necessary with respect to academic articles, when different people within the same field cannot, or do not wish to, cooperate. But generally, the whole point of permitting multiple, competing articles is that this gives a way for very different people to avoid conflict (and thus the need for arbitration). This requires that each creator’s decisions be respected. Still, there would ultimately have to be some reliable means to ensure that truly crankish, idiosyncratic stuff is not added to the main namespace. That is the function of the next item:

(6) A rating system, or in lieu of that, an AI system, may be used for article sorting. An article rating system such as described under Thesis 7 would be an excellent way to determine which articles to show to the end user. If the system has not been built or if there are not enough ratings for a set of articles yet, then an open-source and open-data LLM, set up according to very broad principles of neutrality and fairness, would assign numerical ratings. The ordering of the articles contributed under a topic heading could be partly randomized, on the assumption that the AI ratings are not entirely reliable, and to prevent gaming the system.

(7) Application to search results and hyperlinking. Exactly how the introduction of competing articles affects hyperlinking and search results remains to be determined. Perhaps, if more than one article under a certain topic meets certain objective criteria, they can both be suggested when the user clicks a hyperlink or searches for the topic. It would be best if users could set their own preferences.

Advantages

• GASP preserved for those who care about it. One great advantage of this proposal, from the perspective of current editors, is that it does not require that they change their current practices. They may continue to systematically exclude those who resist the requirements of the GASP framework. They can defend and, in a way, dignify their approach by honestly admitting what their framework has been all along. By supporting other frameworks, they may demonstrate their much-vaunted commitment to diversity, inclusion, and multi-culturalism.
• A revitalized community. There would be large numbers of new participants, from a very wide variety of viewpoints, enthusiastic to get to work. Wikipedia could once again boast of a truly open, welcoming, and global project.
• Richer content. A fascinating body of new content would be created that gives voice to viewpoints currently neglected by Wikipedia’s system. This could become a rich source of comparative cultural studies, useful for students and AI training data alike. Each article would act like an intellectual diplomatic attaché, representative of one framework alongside other frameworks. No longer would Wikipedia present just one Establishment-approved text that represents the “consensus” view of a narrow group of people. Wikipedia would, once again, upset the exclusive prerogatives of the powerful.
• A more accurate representation of diversity. At present, non-Western cultures, minority views, and disfavored ideologies are represented by a single, totalizing, intellectually imperialistic perspective. Permitting multiple, competing articles within various frameworks would give readers a much fuller idea of the actual views held by those operating outside the very narrow, confining bounds of GASP.
• Peace. The atmosphere in general would become more peaceful and collegial. Fundamental editorial conflicts would decline. The possibility of actual decision-making by consensus, albeit consensus within a framework, might be achievable once again.
• Anti-corruption. If done right, this provides a possible solution to the general problem of corruption within Wikipedia, in which the highest bidder pays for a Wikipedia article to read a certain way, to wit: Less corrupt content will always be available.

One of the Nine Theses on Wikipedia series

Footnotes

1. Wikipedia-logo-v2-en.svg by Wikipedia. License: CC BY-SA 4.0, via Wikimedia Commons.[↩]
2. Of course, by this, they must be referring to the original Christianity of the countries from which most English Wikipedia contributors hail. But it is no great stretch to say that not many Wikipedians actually believe the tenets of orthodox Christianity. We will discuss this further presently.[↩]
3. This may come as news to some old Wikipedians who knew me “back in the day.” I converted in 2020 and told my conversion story last winter. [a][↩]
4. Footnotes and links are removed from the following quotation for readability.[↩]
5. Yes, even Muslims, and this matters, because, according to them, Allah is another name given in Arabic to Yahweh, and the origin of his worship was with his revelation to Abram explained in Genesis 12: Muslims agree with Jews and Christians on this. So, Wikipedia’s editors are contradicting religious scholars and rank-and-file believers of all three of these religions.[↩]
6. This may be said to be true even if the believers more often use other names for God.[↩]