Wikipedia:Neutral point of view

Original statement of the policy

A general purpose encyclopedia is a collection of synthesized knowledge presented from a neutral point of view. To whatever extent possible, encyclopedic writing should steer clear of taking any particular stance other than the stance of the neutral point of view.

The neutral point of view attempts to present ideas and facts in such a fashion that both supporters and opponents can agree. Of course, 100% agreement is not possible; there are ideologues in the world who will not concede to any presentation other than a forceful statement of their own point of view. We can only seek a type of writing that is agreeable to essentially rational people who may differ on particular points.

Some examples may help to drive home the point I am trying to make.

1. An encyclopedic article should not argue that corporations are criminals, even if the author believes it to be so. It should instead present the fact that some people believe it, and what their reasons are, and then as well it should present what the other side says.
2. An encyclopedia article should not argue that laissez-faire capitalism is the best social system. (I happen to believe this, by the way.) It should instead present the arguments of the advocates of that point of view, and the arguments of the people who disagree with that point of view.

Perhaps the easiest way to make your writing more encyclopedic, is to write about what people believe, rather than what is so. If this strikes you as somehow subjectivist or collectivist or imperialist, then ask me about it, because I think that you are just mistaken. What people believe is a matter of objective fact, and we can present that quite easily from the neutral point of view. --Jimbo Wales

Introduction

Wikipedia has an important policy: roughly stated, you should write articles without bias, representing all views fairly. This is easily misunderstood. The policy doesn't assume that it's possible to write an article from just one point of view, which would be the one unbiased, "objective" point of view. The Wikipedia neutrality policy says that we should fairly represent all sides of a dispute, and not make an article state, imply, or insinuate that any one side is correct.

It's crucial that we work together to make articles unbiased. It's one of the things that makes Wikipedia work so well.

Writing unbiased text is an art that requires practice.

The following essay explains this policy in depth, and is the result of much discussion. We strongly encourage you to read it, comment on it, and generally come to grips with it.

The basic concept of neutrality

A key Wikipedia policy is that articles should be "unbiased," or written from a "neutral point of view." We use these terms in a precise way that is different from the common understanding. It's crucial to grasp what it means to be neutral (in this sense)--a careful reading of this page will help.

Basically, to write without bias requires that articles do not advocate any specific point of view. Different viewpoints in a controversy are all described fairly. To write articles without bias describes debates rather than arguing one side of the debate.

Why should Wikipedia be unbiased?

Wikipedia is a general encyclopedia, which means it is a representation of human knowledge at some level of generality. But we (humans) disagree about specific cases; for any topic on which there are competing views, each view represents a different theory of what the truth is, and insofar as that view contradicts other views, its adherents believe that the other views are false and therefore not knowledge. Where there is disagreement about what is true, there's disagreement about what constitutes knowledge. Wikipedia works because it's a collaborative effort; but, whilst collaborating, how can we solve the problem of endless "edit wars" in which one person asserts that p, whereupon the next person changes the text so that it asserts that not-p?

A solution is that we accept, for purposes of working on Wikipedia, that "human knowledge" includes all different significant theories on all different topics. So we're committed to the goal of representing human knowledge in that sense. Something like this is surely a well-established sense of the word "knowledge"; in this sense, what is "known" changes constantly with the passage of time, and when we use the word "know" in the sense, we often use so-called scare quotes. In the Middle Ages, we "knew" that demons caused diseases. We now "know" otherwise.

We could sum up human knowledge (in this sense) in a biased way: we'd state a series of theories about topic T, and then claim that the truth about T is such-and-such. But again, consider that Wikipedia is an international, collaborative project. Probably, as we grow, nearly every view on every subject will (eventually) be found among our authors and readers. To avoid endless edit wars, we can agree to present each of these views fairly, and not assert any one of them as correct. That is what makes an article "unbiased" or "neutral" in the sense we are presenting here. To write from a neutral point of view, one presents controversial views without asserting them; to do that, it generally suffices to present competing views in a way that is more or less acceptable to their adherents, and also to attribute the views to their adherents.

To sum up the primary reason for this policy: Wikipedia is an encyclopedia, a compilation of human knowledge. But since Wikipedia is a community-built, international resource, we surely cannot expect our collaborators to agree in all cases, or even in many cases, on what constitutes knowledge in a strict sense. We can, therefore, adopt the looser sense of "human knowledge" according to which a wide variety of conflicting theories constitute what we call "knowledge." We should, both individually and collectively, make an effort to present these conflicting views fairly, without advocating any one of them.

There is another reason to commit ourselves to this policy. Namely, when it is clear to readers that we do not expect them to adopt any particular opinion, this leaves them free to make up their minds for themselves, and thus to encourage in them intellectual independence. Totalitarian governments and dogmatic institutions everywhere might find reason to be opposed to Wikipedia, if we succeed in adhering to our nonbias policy: the presentation of many competing theories on a wide variety of subjects suggests that we, the creators of Wikipedia, trust readers' competence to form their own opinions themselves. Texts that present multiple viewpoints fairly, without demanding that the reader accept any one of them, are liberating. Neutrality subverts dogmatism, and nearly everyone working on Wikipedia can agree this is a good thing.

What is the neutral point of view?

What we mean isn't obvious, and is easily misunderstood.

There are many other possible valid understandings of what "unbiased," "neutral," etc. mean. The notion of "unbiased writing" that informs Wikipedia's policy is "presenting conflicting views without asserting them." This needs further clarification, as follows.

First, and most importantly, consider what it means to say that unbiased writing presents conflicting views without asserting them. Unbiased writing does not present only the most popular view; it does not assert the most popular view as being correct after presenting all views; it does not assert that some sort of intermediate view among the different views is the correct one. Presenting all points of view says, more or less, that p-ists believe that p, and q-ists believe that q, and that's where the debate stands at present. Ideally, presenting all points of view also gives a great deal of background on who believes that p and q and why, and which view is more popular (being careful not to imply that popularity implies correctness). Detailed articles might also contain the mutual evaluations of the p-ists and the q-ists, allowing each side to give its "best shot" at the other, but studiously refraining from saying who won the exchange.

A point here bears elaboration. We said that the neutral point of view is not, contrary to the seeming implication of the phrase, some actual point of view that is "neutral," or "intermediate," among the different positions. That represents a particular understanding of what "neutral point of view" means. The prevailing Wikipedia understanding is that the neutral point of view is not a point of view at all; according to our understanding, when one writes neutrally, one is very careful not to state (or imply or insinuate or subtly massage the reader into believing) that any particular view at all is correct.

Another point bears elaboration as well. Writing unbiasedly can be conceived very well as representing disputes, characterizing them, rather than engaging in them. One can think of unbiased writing as the cold, fair, analytical description of debates. Of course, one might well doubt that this can be done at all without somehow subtly implying or insinuating that one position is correct. But experienced academics, polemical writers, and rhetoricians are well-attuned to bias, both their own and others', so that they can usually spot a description of a debate that tends to favor one side. If they so choose, with some creativity, they can usually remove that bias.

Now an important qualification. Articles that compare views need not give minority views as much or as detailed a description as more popular views. We should not attempt to represent a dispute as if a view held by only a small minority of people deserved as much attention as a majority view. That may be misleading as to the shape of the dispute. If we are to represent the dispute fairly, we should present competing views in proportion to their representation among experts on the subject, or among the concerned parties. None of this, however, is to say that minority views cannot receive as much attention as we can possibly give them on pages specifically devoted to those views. There is no size limit to Wikipedia. But even on such pages, though a view is spelled out possibly in great detail, we still make sure that the view is not represented as the truth.

Bias per se need not be conscious. For example, beginners in a field often fail to realize that what sounds like common sense is actually biased in favor of one particular view. (So we not infrequently need an expert in order to render the article entirely unbiased.) To take another example, writers can, without intent, propagate "geographical" bias, by for example describing a dispute as it is conducted in one country without knowing that the dispute is framed differently elsewhere.

An alternate formulation

We sometimes give an alternative formulation of the nonbias policy: assert facts, including facts about opinions--but don't assert opinions themselves. By "fact," on the one hand, we mean "a piece of information about which there is no serious dispute." In this sense, that a survey produced a certain published result is a fact. That Mars is a planet is a fact. That Socrates was a philosopher is a fact. No one seriously disputes any of these things. So we can feel free to assert as many of them as we can. By "opinion," on the other hand, we mean "a piece of information about which there is some dispute." There's bound to be borderline cases where we're not sure if we should take a particular dispute seriously; but there are many propositions that very clearly express opinions. That God exists is an opinion. That the Beatles were the greatest band is an opinion. That the United States was wrong to drop the atomic bomb over Hiroshima and Nagasaki is an opinion.

For determining whether something is fact or opinion in this sense, it does not matter what the actual truth of the matter is; there can at least in theory be "false facts" (things that everybody agrees upon, but which are, in fact, false), and there are very often "true opinions," though necessarily, it seems, there are more false ones.

Wikipedia is devoted to stating facts and only facts. Where we might want to state opinions, we convert that opinion into a fact by attributing the opinion to someone. So, rather than asserting, "God exists," which is an opinion, we can say, "Most Americans believe that God exists," which is fact, or "Thomas Aquinas believed that God exists," which is also fact. In the first instance we assert an opinion; in the second and third instances we convert that opinion into fact by attributing it to someone. In presenting an opinion, moreover, it is important that we bear in mind that there are sometimes even disagreements about how opinions are best stated; sometimes, it will be necessary to qualify the description of an opinion or to present several formulations, simply to arrive at an solution that fairly represents all the leading views of the situation.

But it's not enough, to express the Wikipedia nonbias policy, just to say that we should state facts and not opinions. When asserting a fact about an opinion, it is important also to assert facts about competing opinions, and to do so without implying that any one of the opinions is correct. It's also generally important to give the facts about the reasons behind the views, and to make it clear who holds them. (It's often best to cite a prominent representative of the view.)

Fairness and sympathetic tone

If we're going to characterize disputes fairly, we should present competing views with a consistently positive, sympathetic tone. A lot of articles end up as partisan commentary even while presenting both points of view; this is wrong. Even when a topic is presented in terms of facts rather than opinion, an article can still radiate an implied stance through either selection of which facts to present, or more subtly their organization--for instance, refuting opposing views as one goes makes them look a lot worse than collecting them in an opinions-of-opponents section.

We should, instead, write articles with the tone that all positions presented are at least plausible. Let's present all competing views sympathetically. We can write with the attitude that such-and-such is a good idea, except that, on the view of some detractors, the supporters of said view overlooked such-and-such a detail. If we can't do that, we will probably write stuff with so much contempt that subsequent edits are going to have a hard time doing anything but veiling it.

Characterizing opinions of people's work

A special case is the expression of aesthetic opinions. Wikipedia articles about art, artists, and other creative topics (e.g., musicians, actors, books, etc.) have tended toward the effusive. This is out of place in an encyclopedia; we might not be able to agree that so-and-so is the greatest guitar player in history. But it is important indeed how some artist or some work has been received by the general public or by prominent experts. Providing an overview of the common interpretations of a creative work, preferably with citations or references to notable individuals holding that interpretation, is appropriate. For instance, that Shakespeare is one of the greatest authors of the English language is a bit of knowledge that one should learn from an encyclopedia. Notice, determining how some artist or work has been received publicly or critically might require research; but that reception, unlike the idiosyncratic opinion of the Wikipedia article writer, is an opinion that really matters.

A consequence: writing for the enemy

Those who constantly attempt to advocate their views on politically charged topics (for example), who seem not to care about whether other points of view are represented fairly, are violating the nonbias policy ("write unbiasedly"). But the policy entails that it is our job to speak for the other side, and not just represent our own views. If we don't commit ourselves to doing that, Wikipedia will be much, much weaker for it. We should all be engaged in explaining each other's points of view as sympathetically as possible.

In saying this, we are spelling out what might have been obvious from an initial reading of the policy. If each of us is permitted to contribute biased stuff, then how is it possible that the policy is ever violated? The policy says, "Go thou and write unbiasedly". If that doesn't entail that each of us should fairly represent views with which we disagree, then what does it mean? Maybe you think it means, "Represent your own view fairly, and let others have a say." But consider, if we each take responsibility for the entire article when we hit "save", then when we make a change that represents our own views but not contrary views, or represents contrary views unfairly or incompletely, surely we are adding bias to Wikipedia. Does it make sense not to take responsibility for the entire article? Does it make sense to take sentences and say, "These are mine"? Perhaps, but in a project that is so strongly and explicitly committed to neutrality, that attitude seems out of place.

The other side might very well find your attempts to characterize their views substandard, but it's the thought that counts. In resolving disputes over neutrality issues, it's far better that we acknowledge that all sides must be presented fairly, and make at least a college try at presenting the other sides fairly. That will be appreciated much more than not trying at all.

"Writing for the enemy" might make it seem as if we were adding deliberately flawed arguments to Wikipedia, which would be a very strange thing to do. But it's better to view this (otherwise puzzling) behavior as adding the best (published) arguments of the opposition, preferably citing some prominent person who has actually made the argument in the form in which you present it, stating them as sympathetically as possible. Academics, e.g., philosophers, do this all the time.

An example

It might help to consider an example of a biased text and how Wikipedians have rendered it at least relatively unbiased.

On the abortion page, early in 2001, some advocates had used the page to exchange barbs, being unable to agree about what arguments should be on the page and how the competing positions should be represented. What was needed--and what was added--was an in-depth discussion of the different positions about the moral and legal viability of abortion at different times. This discussion of the positions was carefully crafted so as not to favor any one of the positions outlined. This made it rather easier to organize and understand the arguments surrounding the topic of abortion, which were each then presented sympathetically, each with its strengths and weaknesses.

There are numerous other "success stories" of articles that began life as virtual partisan screeds but were nicely cleaned up by people who concerned themselves with representing all views clearly and sympathetically.

Another example

Karada offered the following excellent advice in the context of the Saddam Hussein article:

You won't even need to say he was evil. That's why the article on Hitler does not start with "Hitler was a bad man" -- we don't need to, his deeds convict him a thousand times over. We just list the facts of the Holocaust dispassionately, and the voices of the dead cry out afresh in a way that makes name-calling both pointless and unnecessary. Please do the same: list Saddam's crimes, cite your sources.

Objections and clarifications

What follows is a list of common objections, or questions, regarding Wikipedia's nonbias policy, followed by replies.

There's no such thing as objectivity

Everybody with any philosophical sophistication knows that. So how can we take the "neutrality" policy seriously? Neutrality, lack of bias, isn't possible.

This is probably the most common objection to the neutrality policy. It also reflects the most common misunderstanding of the policy. The misunderstanding is that the policy says something about the possibility of objectivity. It simply does not. In particular, the policy does not say that there even is such a thing as objectivity, a "view from nowhere" (in Thomas Nagel's phrase)--such that articles written from that point of view are consequently objectively true. That isn't the policy and it is not our aim! Rather, we employ a different understanding of "neutral" and "unbiased" than many might be used to. The policy is simply that we should characterize disputes rather than engage in them. To say this is not to say anything contentious, from a philosophical point of view; indeed, this is something that philosophers are doing all the time. Sophisticated relativists will immediately recognize that the policy is perfectly consistent with their relativism.

If there's anything possibly contentious about the policy along these lines, it is the implication that it is possible to characterize disputes fairly, so that all the major participants will be able to look at the resulting text, agreeing that their views are presented sympathetically and as completely as possible (within the context of the discussion). It is an empirical question, not a philosophical one, whether this is possible; and that such a thing is indeed possible is evident simply by observing that such texts are being written daily by the most capable academics, encyclopedists, textbook writers, and journalists.

Pseudoscience

How are we to write articles about pseudoscientific topics, about which majority scientific opinion is that the pseudoscientific opinion is not credible and doesn't even really deserve serious mention?

If we're going to represent the sum total of "human knowledge"--of what we believe we know, essentially--then we must concede that we will be describing views repugnant to us without asserting that they are false. Things are not, however, as bad as that sounds. The task before us is not to describe disputes fairly, on some bogus view of fairness that would have us describe pseudoscience as if were on a par with science; rather, the task is to represent the majority (scientific) view as the majority view and the minority (sometimes pseudoscientific) view as the minority view, and, moreover, to explain how scientists have received pseudoscientific theories. This is all in the purview of the task of describing a dispute fairly.

There is a minority of Wikipedians who feel so strongly about this problem, however, that they believe Wikipedia should adopt a "scientific point of view" rather than a "neutral point of view." What these people have failed to establish, however, is that there is really a need for such a policy, given that the scientists' view of pseudoscience can be clearly, fully, and fairly explained to those who might be misled by pseudoscience.

Morally offensive views

What about views that are morally offensive to most Westerners, such as racism, sexism, and Holocaust denial, that some people actually have? Surely we are not to be neutral about them?

We can certainly include long discussions that present our moral repugnance to such things; in doing so, we can maintain a healthy, consistent support for the neutral point of view by attributing the view to prominent representatives or to some group of people. Others will be able to make up their own minds and, being reasonable, surely come around to our view. Those who harbor racism, sexism, etc., will surely not be convinced to change their views based on a biased article, which only puts them on the defensive; on the other hand, if we make a concerted effort to apply our nonbias policy consistently, we might give those with morally repugnant beliefs insight that will change those views.

Giving "equal validity"

But wait. I find the optimism about science vs. pseudo-science to be baseless. History has shown that pseudo-science can beat out facts, as those who rely on pseudo-science use lies, slander, innuendo and numerical majorities of its followers to force their views on anyone they can. If this project gives equal validity to those who literally claim that the Earth is flat, or those who claim that the Holocaust never occurred, the result is that it will (inadvertently) legitimize and help promote that which only can be termed evil.

Please be clear on one thing: the Wikipedia neutrality policy certainly does not state, or imply, that we must "give equal validity" to completely repugnant views. It does state that we must not take a stand on them qua encyclopedia writers; but that does not stop us from representing the majority views as such; from fairly explaining the strong arguments against the repugnant views; from describing the strong moral repugnance that many decent people feel toward them; and so forth.

Hence, on the one hand, Wikipedia does not officially take a stand even on such obvious issues, but on the other, it will not look as though we (the authors of Wikipedia) had accorded equal credibility to morally repugnant views. Given that the authors of Wikipedia represent a rough cross-section of the educated public, our readers can expect us to have a similar cross-section of opinion about extremism: most of us abhor it.

Americo-centric point of view

Wikipedia seems to have an Americo-centric point of view. Isn't this contrary to the neutral point of view?

Yes, it certainly is, and it has no defenders on Wikipedia. The presence of articles written from an exclusively United States point of view is merely a reflection of the fact that there are many Americans working on the project, which in turn is merely a reflection of the fact that so many Americans are online and working on the English project/

This is an ongoing problem that can be corrected by active collaboration from people outside of the U.S., of whom there are many.

This is not only a problem in English. The French Wikipedia may reflect a Continental French bias, the Japanese Wikipedia suffers from Japanese bias, and so on.

Lack of neutrality as an excuse to delete

The neutrality policy is used sometimes as an excuse to delete texts that are perceived as biased. Isn't this a problem?

In many cases, yes. Most of us believe that the mere fact that some text is biased is not enough, by itself, to delete the text outright. If it contains perfectly valid information, the text should simply be edited accordingly, and certainly not deleted.

There's sometimes trouble determining whether some claim is true or useful, particularly when there are few people on board who know about the topic. In such a case, it's a good idea to raise objections on a talk page; if one has some reason to believe that the author of the biased material will not be induced to change it, we have sometimes taken to removing the text to the talk page itself (but certainly not deleting it entirely). But the latter should be done more or less as a last resort, never merely as a way of punishing people who have written something biased.

Dealing with biased contributors

I agree with the nonbias policy but there are some here who seem completely, irremediably biased. I have to go around and clean up after them. What do I do?

Unless the case is really egregious, maybe the best thing is to call attention to the problem publicly, pointing the perpetrators to this page (but politely--one gets more flies with honey) and asking others to help. If the problem is really serious, Jimbo Wales might be enlisted to beat the person over the head (so to speak) and, in the most recalcitrant cases, ask them to leave the project. There must surely be a point beyond which our very strong interest in being a completely open project is trumped by the interest the vast majority of our writers have, in being able to get work done without constantly having to fix the intrusions of people who do not respect our policy.

Avoiding constant disputes

How can we avoid constant and endless warfare over neutrality issues?

Would that people asked this question more often. We should never debate about how Wikipedia should be biased. It shouldn't be biased at all.

The best way to avoid warfare over bias is to remember that we are all reasonably intelligent, articulate people here, or we wouldn't be working on this and caring so much about it. We have to make it our goal to understand each others' perspectives and to work hard to make sure that those other perspectives are fairly represented. When any dispute arises as to what the article "should" say or what is "true," we must not adopt an adversarial stance; we must do our best to step back and ask ourselves, "How can this dispute be fairly characterized?" This has to be asked repeatedly as each new controversial point is stated. It is not our job to edit Wikipedia so that it reflects our own idiosyncratic views and then defend those edits against all comers; it is our job to work together, mainly adding new content, but also, when necessary, coming to a compromise about how a controversy should be described, so that it is fair to all sides.

Making necessary assumptions

What about the case where, in order to write any of a long series of articles on some general subject, we must make some controversial assumptions? That's the case, e.g., in writing about evolution. Surely we won't have to hash out the evolution-vs.-creationism debate on every such page?

No, surely not. There are virtually no topics that could not proceed without making some assumptions that someone would find controversial. This is true not only in evolutionary biology, but also philosophy, history, physics, etc.

It is difficult to draw up general principles on which to rule in specific cases, but the following might help: there is probably not a good reason to discuss some assumption on a given page, if an assumption is best discussed in depth on some other page. Some brief, unobtrusive pointer might be apropos, however. E.g., in an article about the evolutionary development of horses, we might have one brief sentence to the effect that some creationists do not believe that horses (or any other animals) underwent any evolution, and point the reader to the relevant article. If there is much specific argumentation on some particular point, it might be placed on a special page of its own.

Concerns about "writing for the enemy"

I'm not convinced by what you say about "writing for the enemy." I don't want to write for the enemy. Most of them rely on stating as fact many things which are demonstrably false. Are you saying that, to be neutral in writing an article, I must lie, in order to faithfully represent the view I disagree with?

This is a misunderstanding what the neutrality policy says. You aren't claiming anything, except to say, "So-and-so argues that such-and-such, twiddle dee dee, and therefore, QED." This can be done with a straight face, with no moral compunctions, because you are attributing the claim to someone else. That's the important thing here! If we are summing up human knowledge on a subject, in the sense above-defined, then you are leaving out important information when you omit so-and-so's argument.

It's worth observing that some scholars are trained so that, even when trying to prove a point, one brings forth counter-arguments so that one can explain why the counter-arguments fail. Such training also gives one a better knowledge of source material and what may have been rejected over the years. Something very much like the neutral point of view is an assumption among scholars--if it isn't adhered to, or if only those facts that prove a particular point are used, one might lose one's reputation.

Other objections

I have some other objection. Where should I ask it?

Before asking it, please review the links below. Many issues surrounding the neutrality policy have been covered before very extensively. If you have some new contribution to make to the debate, you could try Talk:Neutral_point_of_view, or bring it up on the Wikipedia-l mailing list.

Other resources

• The Nupedia policy on lack of bias
• Meta-Wikipedia draft and commentary
• Examples
• Examples Debate
• Natural point of view
• Understand Bias
• Most controversial subjects in wikipedia
• Words that should not be used in wikipedia articles
• Talk:Creationism
• m:Positive tone

External links

• AssumeGoodFaith and
• NeutralPointOfView, both on MeatballWiki

[[fr:Wikip%E9dia:La_neutralit%E9_de_point_de_vue]]