On neutrality

The original article can be found at larrysanger.org.

---

As a teenager, I habitually scanned encyclopedias, newspapers, and textbooks—angrily. I was on the lookout for bias. I had discovered, to my frequent irritation, that writers of authoritative texts would present only one side of a dispute, as if it were uncontroversially true. Honest and important debates were treated as if they didn’t exist. An encyclopedia writer, for example, was supposed to be an objective source imparting facts. Where a dispute existed, the writer was supposed to represent a broad range of opinions fairly, and anything controversial should have been left up to me to decide for myself. Instead, these writers were indoctrinating me, or trying to. What an abuse of a position of authority! How infuriating!

Later in life, given the opportunity to start an encyclopedia, I was a zealot for neutrality. My teenage ire at shamefully biased writers and editors found expression in Nupedia’s neutrality policy, which in turn became Wikipedia’s. I defended these policies at length against criticisms, but I never carefully articulated or defended my views on neutrality in a paper. Finally, I have an opportunity to do so. Ballotpedia has asked me to write this paper as part of their effort to adopt more formal, systematic principles and practices surrounding neutrality.

In this essay, I will defend neutrality as the preferred policy for some types of writing. I will define neutral writing, lay out four arguments in favor of a policy neutral writing, and refute some common criticisms. I will conclude by suggesting that publishers adopt both neutrality guidelines and editorial processes to make sure that the guidelines are followed.

Criticisms are lurking in the background throughout this essay, and let me tip my hat slightly at them before I dive in.

I concede that neutrality is a headache. It is so difficult to achieve that bias is the norm, and it will continue to be the norm unless it is deliberately eradicated.

Neutrality, even if it is a writer’s aim, can be very hard to achieve. Neutral writers have to have mastered their subject and then be extremely careful and fair-minded. Otherwise, they are almost sure to display bias. In my experience, neutrality doesn’t come easily to any of us, but is a discipline that must be practiced. Some of us don’t even seem to have the concept, or rather we confuse it with objectivity or being scientific or describing a middle ground or taking the most mainstream position. Even if we understand the ideal of neutrality—understanding it does make it easier to achieve—and our writing is successfully unbiased, others may misunderstand what we have achieved. Partisans might think we are soft on “the other side” (meaning the one they’re opposed to) or even biased in its favor. Neutral writers might even be accused, perversely, I think, of “false balance.” So, as difficult as it is, a neutrality policy might seem more trouble than it’s worth.

Besides, not everyone is on board with neutrality. Bias (in the sense of tendentious, one-sided communication) has its apologists. Some claim that neutrality is not just difficult but actually impossible. We are each hopelessly biased and we simply cannot keep our biases from coming out, they say. Some go further to claim that neutrality is simply wrong or unreasonable in many cases. Some points of view do not deserve any expression. And where the truth is known, it is wrong to pretend it is not.

I disagree. What if I were to tell you that, if you write biased encyclopedia and newspaper articles, you’re doing a moral wrong, and what you’re doing is on a par with propaganda? Ridiculous, you say? But it’s true, as I’ll argue further down.

What is neutrality?

1. Defining “neutrality”

First, let’s define neutrality.

Here’s the basic idea: if you’re neutral, you don’t take a position. You present all sides fairly and let your reader decide which is correct.

A disputed topic is treated neutrally if each viewpoint about it is not asserted but rather presented (1) as sympathetically as possible, bearing in mind that other, competing views must be represented as well, and (2) with an equitable amount of space being allotted to each, whatever that might be.

On this account, neutrality is a concept dealing specifically with disputed topics, and it has three basic requirements.

First, if an issue that is mentioned in the text is disputed, the text takes no position on the issue. Neutrality is not some midpoint in between competing options. A political moderate’s positions are not “the neutral positions”: they are positions as well. Neutral writing takes no position, left, right, or middle.

Second, there’s the requirement of tone, or the strength of the case made for a viewpoint. Basically, if you’re going to be neutral, you have to represent all the main views about the topic, and you have to represent them all sympathetically, i.e., according to their best, most convincing arguments, evidence, and representatives.

Third, there’s the question of how much space it is fair or equitable to spend in a text on the different sides. Prima facie, it would seem that spending a numerically equal amount of space on both (or all) sides is fair, but it doesn’t always work out that way. Exactly how to apportion limited space is a complex question I’ll address further down.

Basically, to write neutrally is to lay out all sides of any disputed question, without asserting, implying, or insinuating that any one side is correct. If a debated point is mentioned, you represent the state of the debate rather than taking a stand.

As I will explain below, it is probably practically impossible to achieve perfect neutrality, if that were understood to mean neutrality with respect to all cultures and all historical eras. I will also mention the notion of a “good enough” neutrality. One observer noticed that these concessions seem to commit me to the proposition that neutrality comes in gradations or aspects—which is something I happily admit. What I advocate might be described as a strict or professional standard of neutrality.

2. Some principles of neutrality

Here are some general principles that are more or less implied by this definition. I don’t claim that these principles have no exceptions but only that they give a fairly good idea of what neutrality entails.

It is impossible to tell reliably what side of a disputed issue the author of the text is on, if the text is neutral with respect to that issue. The text avoids word quantity, choice, and tone that favors one side over another. Both or all sides agree that their side is fairly represented. Barring that, the text will tend to anger or dissatisfy everyone equally, although for different reasons. Generally, there is a focus on or preference for agreed-upon “facts.” If an opinion is included, it is attributed to a source. The debate is not engaged but rather described and characterized, including information about proportions of people on the different sides, where available. Controversial claims—i.e., claims that a party to a dispute might want to take issue with—are all attributed to specific sources. The author does not personally assert such claims. Biased sources are either eschewed or used in approximately equal numbers on both or all sides throughout a document. A document that uses many biased sources on only one side looks biased itself. When there is a “significant” (this word admittedly glosses over an important problem) ongoing debate and a source implies a definite stand on it in an article that is not about that debate, at the very least there must be some acknowledgment in the text that a disagreement exists. When it makes no sense for articles to be individually neutral, reporting and publications that are neutral with respect to a debate, or a field, will publish in equal amounts on both or all sides of an issue. If a publication favors one side because more papers are received on that side, or because more of the research community embraces that side, that might appear fair and reasonable, but it is not neutral and equally balanced: it will tend to make one side look better than the other.

3. Neutrality distinguished from rationality, scientific writing, and objectivity

Let’s get clear about one thing: neutrality is different from being rational, scientific, or objective. People often use these terms as if they were synonyms. For example, at least one journalist has argued that “ ‘neutral’ journalism can’t die fast enough,” but only by glossing over the distinction between neutrality and objectivity, and his piece simply did not consider the notion of neutrality discussed in this paper.

Rationality means (roughly) following the rules of logic and careful observation and proportioning belief to the amount and quality of the evidence. Being scientific (also roughly) means following the scientific method and, again, not believing a hypothesis unless it has been well established by the canons of science. Objectivity means deciding what to believe (or write) not based on mere feelings or prejudice, but strictly based on substantive considerations of evidence that are supposedly independent of our personal views.

Rationality, scientific method, and objectivity all concern epistemology or methodology, i.e., they are standards of knowledge, or else methods we use to determine our beliefs in an attempt to arrive at knowledge. Neutrality is in a different category altogether: it’s a standard of exposition, of expression. If I write neutrally, I am simply refusing to take a position. As I am neither evaluating or staking out a claim, the rationality, scientific merit, and objectivity of my claim are not at issue. It is true that in writing neutrally, I might help others to be objective. But neutrality is a style of writing, while objectivity is a state of mind. Expression is neutral; people and their mental states are (ideally) objective.

Neutrality and those other qualities are orthogonal. You can have one without the other. In fact, if you have written much in a neutral way, then you’ve probably represented views fairly that you believe to be irrational, unscientific, or subjective. You have simply not taken a position. Similarly, you can write a perfectly rational, scientific, and objective treatment of a topic—which takes one particular position in a hotly-debated dispute.

For example, suppose in a piece of writing, you discuss the measles vaccination controversy. If you’re going to write in accordance with the canons of science (and also rationality and objectivity, no doubt), then in my opinion, you will support the view that children should be vaccinated, period. But if what you write is going to be neutral with respect to current American society, you will have to withhold any such forthright claim. Rather, you’ll report that while a very large majority of doctors and scientists strongly advise vaccination, something like 9% of the population thinks that the measles vaccine is unsafe. You will not take or project any stand one way or another, although, to be fair to the vast majority of the public in favor of measles vaccination, in some kinds of articles you might spend only about 9% of the article on the anti-vaccination position.

4. Does equitable division of space always mean equal space?

You might well disagree with the latter sentence. Even if it is a minority view, why should the anti-vaccination position get only 9% of an article? It does not seem obvious that that’s fair. If we’re trying to help people make up their minds, why isn’t 50% fair and neutral? If neutrality means, ultimately, that the reader cannot guess your position as the writer, then a case can be made for 9%, and a case can be made for 50%.

I suggest that it depends on the type of writing. Consider two examples.

First, news reporting of scientific studies: each time a news service reports a new study on a scientific controversy, neutrality does not require that the journalist give equal space to all competing theories. That would be unreasonable. Neutrality, then, is not an article-level concept but a service-level concept. For example, the news service might have a rule to the effect that major studies supporting competing theories should be covered in proportion to how often they appear in major journals. Perhaps the journalist is obligated to acknowledge briefly that there are one or more competing theories, and if there happens to be a debate or disagreement about the study itself, that debate should be covered neutrally.

Second, articles introducing policy debates: surely we do expect an encyclopedia or debate website article on the vaccination issue to be scrupulously even-handed, and perhaps that really does mean giving approximately equal time to both or all sides. But why? Why should a description of the debate give 50% of the space to the anti-vaccination side if only 9% of the public (and probably less than 1% of doctors and medical researchers) are on that side? Why not give something like 9% to the anti-vaccination side? Wouldn’t that be fair? Why should the minority position be dignified to that extent?

I think the answer stems from the premise that the major purpose of reporting on a controversy is to enable people to come to their own conclusions about it. In that case, devoting more space to one side, even if it is the majority side, would suggest that that side is correct. This in turn would interfere with the mission of supporting readers in deciding the issue for themselves. Scrupulously presenting both or all sides in approximately equal proportions, and with the best available arguments, etc., gives readers no clues as to “the right” answer, requiring them to fall back on their own critical thinking skills. In other words, covering the debate in a strictly even-handed way enables readers to be more rational.

Advocacy journalism, I am well aware, rejects my premise that the purpose of reporting on a controversy is to enable people to come to their own conclusions about it, though. Still, that ought to clarify at least somewhat better what I mean by the third part, on fair apportionment of space, of my definition of “neutrality.” It gets more complicated, as I will explain in sections toward the end of this paper.

5. Traditionally neutral texts

There are at least three categories of text that are traditionally neutral: encyclopedia and reference writing, straight news writing, and basic textbooks. These aren’t always neutral, of course, but they are generally expected to be. Why? Why should they be expected to be neutral when opinion pieces, art, documentaries, scholarly papers and monographs, and various other forms aren’t?

Lower-level textbooks (through the junior high school level, say) are expected to impart facts and avoid controversy in order to give a student a basic foundation of knowledge. If an elementary school textbook is full of opinion, theory, and controversy, it’s less likely to be used, because students are expected to begin their studies with commonly known facts, which are more than enough to learn. At higher levels, a range of opinions and controversial theories may and indeed should be included, but even then, if they are not neutrally presented, the text runs the risk of alienating students and their parents. By making textbooks neutral, we make them satisfactory for everyone in a diverse society. Of course, there are biased texts—especially at the high school and college levels. Thoroughly biased texts, which inculcate only one political or religious view, are used especially by highly ideological teachers and professors, at religious schools, and in home schools; but such texts and teaching generally earn the contempt of many of us because the students who emerge will not be so well prepared for life in a more intellectually diverse society. Besides all that, texts that are biased are more likely to get the facts wrong, as I will argue with respect to journalism.

Not all journalism is or is expected to be neutral, of course. We think of journalism as straight reporting, but it also includes outright opinion pieces, advocacy journalism, and opinion and debate broadcasts. These biased forms have become dominant in 21st century society, especially on cable news shows, opinion websites, blogs, most news magazines, and talk radio. Yet even in our modern, cynical news consumption societies, “straight news” is still a fairly well-understood category and is generally expected to be neutral, even if it comes from Fox News or MSNBC. In a straight news story, the journalist’s opinion is irrelevant and distracting. Partly this is because there is typically a lot of news to report and reader time is quite limited as well. In addition, there are a lot of people—I am one—who simply don’t like any opinion mixed with their straight facts.

But a deeper reason is that ideologues seem to get the facts wrong so often when reporting the news. Reporting the straight facts speedily and readably with a maximum of accuracy and relevance is surprisingly difficult. Seeing the world always through red- or blue-colored lenses makes writing easier, but the hard job of accuracy on a deadline harder: it tends to blind writers to facts that sit poorly with their filters. So we naturally and rightly distrust the factual claims contained in reporting that strikes us as biased (cf. the “backfire effect”).

Finally, the function of encyclopedias and other reference material is to serve as highly relevant compendia of searchable facts. They are expected to be neutral for similar reasons to the foregoing. On the one hand, mixing opinion with straight factual writing wastes our time and distracts us from what we’re using reference materials for: learning the basic facts of what is known and believed about a topic. On the other hand, again, we don’t trust a reference to catalog complex facts reliably if it is also trying to persuade us of a particular point of view. Getting the facts right is hard enough as it is.

So, for the rest of this essay, I’ll refer to “traditionally neutral texts,” meaning encyclopedias, straight news, and textbooks.

By the way, there have been societies in which the news, encyclopedias, and textbooks were, all three, expected to reflect a single point of view, as if none other were possible: totalitarian societies are like that. The old Soviet Union—with its Pravda newspaper serving as the voice of the Communist Party, its Great Soviet Encyclopedia the source of state-approved facts, and its notoriously propaganda-laden textbooks—is perhaps the most famous historical example. Such openly, complacently biased publications strike us in the 21st century West as deeply problematic subversions of their traditional forms, serving as especially pungent illustrations of why totalitarian societies are so problematic.

Four arguments for neutrality

6. Neutrality respects personal autonomy

So, why neutrality?

I have attempted to explain why, traditionally, we expect certain texts to be neutral. Next, I will advance four arguments for neutrality that apply equally to all traditionally neutral texts.

My first argument is the longest and most important, and it’ll take several sections to go through.

Consider some examples of neutrality and objectivity:

• a science textbook fairly explaining competing theories
• a detailed encyclopedia article breaking down competing narratives and explanations of the Great Depression, and you can’t tell what side the author is on
• a scrupulously balanced news article clarifying the latest policy debate, with all sides fully and sympathetically presented

I admire such approaches. Probably, you do too. But why do we?

It’s because neutral writing respects the reader. It shows the author is treating us like adults who wish to make up our minds rationally, using reason, logic, and evidence. If somebody is being conspicuously neutral, it inspires confidence that we are getting the facts, all the relevant facts, with emotion-driven rhetoric left out. We’re being left to decide the matter for ourselves, rationally.

Someone who writes neutrally supports our natural desire to be rational and thus puts us on the road to truth. The choice of belief is left entirely up to us.

Neutrality, I’ll argue, respects our personal autonomy.

Autonomy can be roughly defined as the capacity to govern oneself freely, independently of other influences generally (metaphysical autonomy), of the moral dictates of others (moral autonomy), of law and government control (political autonomy), of religious dictates (religious autonomy), and of the pressure and indoctrination of authorities such as educators (intellectual autonomy).

It is our intellectual autonomy that neutral encyclopedists, journalists, and educators respect. They leave us free to make up our own minds for ourselves. Naturally, we appreciate that—assuming we value our autonomy, as we should and typically do.

Propagandists, by contrast, sometimes insist that neutrality is impossible or wrong, and they use such assertions as an excuse for taking definite, controversial positions in traditionally neutral forms. Hacks try to control their readers. They don’t want to leave them free to make up their own minds. They don’t respect their autonomy. Propagandists aren’t interested in giving readers the tools they need to decide rationally, for themselves; they want to indoctrinate or trick them into believing precisely the way they believe.

7. Independent, rational deliberation respects autonomy

In my mind, autonomy is bound up with two good things: free will and the Enlightenment.

Autonomy is essentially the same as freedom—as in the freedom of the will—and therefore it is deeply important to morality. My view is that to act with free will is simply to act with an unencumbered, mature ability to think our decisions through. Our ability to deliberate rationally on what we ought to do is what gives us our freedom, or autonomy. It is also what gives us our dignity as individuals: it is our own intellects—our independent, reflective minds and our ability to make them up for ourselves in a mature, adult fashion—that command that basic level of respect we call dignity.

Now, our beliefs themselves are frequently out of our direct control. For example, if I am an atheist, I cannot simply up and decide, “Today I’m going to believe in God.” But we can control the inputs of our beliefs. We can control whether our beliefs are informed by facts and reasoning, or instead by emotion and rhetoric. We can control how long and how carefully we think a view through, before we accept it or that we are committed to it.

If we carefully think through the issues, especially if we consider all sides and all the evidence, then we embrace, as genuinely our own, whatever conclusion we come to. That we have reasoned our way to a conclusion means we accepted it freely. By contrast, if we simply find ourselves with an opinion after idly, passively receiving messages from our friends and from mass media, we are less likely to take responsibility for that view. It seems to be less “our own” and easier to reconsider.

If that’s true, then the more independently deliberative we are, over the course of our lives, the freer we are. This is the conclusion that Spinoza came to, famously, in his Ethics (e.g., see this). The more that we guide our beliefs by our own careful reasoning and observation, the more freedom or autonomy we have.

Intellectual autonomy was naturally a key feature of the Enlightenment, and it remains robust and important to this day; we’re still encouraged to think for ourselves far more than we were in the Middle Ages. Immanuel Kant opened his essay, “What is Enlightenment?” with these ringing words: Enlightenment is man’s emergence from his self-imposed nonage [immaturity]. Nonage is the inability to use one’s own understanding without another’s guidance. This nonage is self-imposed if its cause lies not in lack of understanding but in indecision and lack of courage to use one’s own mind without another’s guidance. Dare to know! (Sapere aude.) “Have the courage to use your own understanding,” is therefore the motto of the enlightenment.

We like and admire neutral writing because it fosters our personal enlightenment. Bias, by contrast, treats us like immature children—who are in “nonage”—who cannot be trusted to arrive independently at the beliefs some authority wishes us to have.

8. Neutrality fosters autonomy

I say neutrality respects autonomy. It also fosters autonomy. So, how does it, exactly?

Neutrality has at least two features: (1) it involves presenting competing positions, and (2) it presents them sympathetically, according to their strongest arguments. Let’s take these in turn.

(1) If various competing views are presented, then we must naturally fall back on our own resources. Neutral writing removes the crutch of simply accepting the single view that the author informs us of. If we are simply “told the One Truth” about a topic, then we do not engage our own brains or curiosity, and we become dependent on the author and less autonomous or free in how we think about the topic. By contrast, if we are given several options, our natural curiosity and desire to settle upon the truth will impel us to reflect on those different options. That is just what intellectual autonomy requires of us.

(2) Moreover, if the competing views are all presented sympathetically, in their strongest forms, we will be given the best data, evidence, and arguments—the tools needed to make up our minds rationally. Armed with those tools, we will be more likely to deliberate rationally in an attempt to arrive at the truth.

This, again, is exactly what intellectual autonomy requires of us: fully adult, intellectual freedom isn’t simply a matter of choosing a view by whim, emotion, or social pressure, but only after carefully examining and comparing the competing arguments and evidence for ourselves.

Bias, by contrast, frequently encourages us simply to leap to a view based on our prejudices, on emotional and social appeals, without thinking the matter through. Even when one side is presented in great rational detail and the other is not mentioned, the lack of mention itself implies without argument that the view is not worth spending any time on. In any event, we are made dependent on whoever propagates a single view. To the extent that we are subject to emotional manipulation or other kinds of fallacy, we are less autonomous or free in how we think about the topic. In short, bias is a tool of control, especially but not only emotional control.

Bias more generally dulls the brain. In an educational setting, it actually discourages people from forming, and exercising, the habits that constitute intellectual autonomy. In civil society generally, it dulls the practice of individual reason, encourages groupthink, and devalues the coin of rational debate.

To sum up both points and to put it metaphorically, neutrality does not give us a free ride. It throws us into the issues and requires us to swim through them under our own power. This can be difficult and frightening (thus Kant’s injunction, sapere aude) but it also makes us feel empowered to decide for ourselves. Neutrality supports us both intellectually and emotionally in the act of exercising autonomous judgment by presenting us with all the options and providing us the tools to judge among them for ourselves.

9. Bias makes us less morally responsible

A case can be made, though I can’t make it here, that there is no more to autonomy than doxastic (belief-related) autonomy. In other words, the only real sort of free will we have stems from our ability to deliberate, to think things through for ourselves. That’s what Spinoza argued in the Ethics, and (after long reflection, of course) I agree with him. But free will is what makes us morally responsible; if we aren’t free when we act, we aren’t responsible for what we do. If that’s true, a surprising consequence follows: bias has a tendency to make us less morally responsible. The more that we are simply told what to think, the less our beliefs are our own, and the less morally responsible we are for our own actions.

Perhaps that just sounds strange. But on reflection, we can see that it is true. Consider the infamous Milgram experiment, in which a browbeating scientist presented his subjects with exactly one option: to deliver what they believed to be (but in fact were not) painful electric shocks to other subjects. There was a doubly shocking bias at work—in favor of continuing the experiment. Think of any number of cults in which the leadership, possessing ultimate religious, moral, and intellectual authority, issues not-to-be-questioned rules and commands, impelling members to do things they would otherwise never dream of doing. The leaders’ injunctions were presented in a thoroughly biased fashion. Reflect on how powerful Nazi and Communist propaganda campaigns made it possible for the hapless citizenry to demonize the “enemies of the state,” dissociate themselves from their own appalling behavior, and act as informants and tools of the state. These are all particularly appalling examples of bias.

If bias can have such horrifying effects on our moral autonomy, then our obligation to strive for neutrality is very strong, indeed. Neutral writing makes us more reflective and therefore more capable of taking moral responsibility for our commitments.

10. Bias as a moral failing

If I am right, neutrality is not just being kind and respectful to readers; it is a positive obligation. When we write neutrally, we help others to be free, and so neutrality is a virtue. Bias, by contrast, is revealed as a moral failing.

Perhaps it should not be surprising that Kant—champion of the Enlightenment—also believed we have a fundamental obligation to respect others’ dignity, their basic, irreducible value as human beings. We should treat others as ends in themselves, Kant famously wrote in his Foundations of the Metaphysics of Morals, not as mere means to our own ends. And what is it that gives us our dignity as human beings? Again, I, like Kant and other philosophers, maintain that it is our autonomy or free will, our ability to deliberate rationally, as adults should, that gives us this basic value, this right to be considered not as a mere means to others’ ends. (Cf. the Second and Third Formulations of the Categorical Imperative.)

When you write with bias, you are treating your readers as your pawns, as mere means to your ends. You are not treating them as autonomous agents, capable of making up their own minds rationally. You are not respecting their dignity.

This is especially the case when you are writing in a traditionally neutral form, one in which people expect to learn the whole story or debate. Admittedly, if you are writing an opinion piece, it is up to the reader to find alternative points of view; while you should still be fair-minded, it is not necessarily your job to make your opponents’ case for them. But if you are writing in a traditionally neutral form, then your audience expects you to lay out everything relevant. If there is some dispute in the relevant community, and you take sides, you’re putting one over on your audience, for your own purposes. I say—though it might sound like hyperbole—that’s morally wrong.

I mean it. But let me clarify. When I say you’re treating your readers as pawns or as mere means to your ends, I don’t mean that you think that way consciously. Maybe what I’m suggesting never occurred to you. But the fact of the matter is that, if you’re writing in a traditionally neutral form, and if you give your readers just one out of various possible views to consider, or only one is presented sympathetically or with all the most convincing evidence and argument, then you are in fact treating your readers fraudulently: you are falsely portraying a controversial issue as if there were only one view that an uninformed person would want to be fully and fairly informed about. Neutrality respects people as autonomous individuals, capable of hearing all sides and judging the evidence fairly for themselves.

Encyclopedists, journalists, and textbook writers should all take note: neutrality is the best policy for free people.

11. Neutrality as adversarial, bias as inquisitorial

The phraseology “hearing all sides and judging the evidence fairly” suggests an analogy: an encyclopedist acts as a fair and impartial judge, enabling the dueling sides to call their witnesses and present their cases as forcefully as possible. Presented with a full airing of all evidence that both sides deem relevant, the jury will, we hope, arrive at a just verdict.

In an analogy with neutral writing, the judge, who organizes and oversees the proceedings, is the author. The counsel are the partisans on the sides of some issue discussed, the witnesses are experts and eyewitnesses, and the evidence and arguments are precisely analogous in both cases. The jury is the reader; the jury’s verdict, the reader’s judgment of the truth. The analogy, then, is that just as we trust that a fair presentation of both sides will result in a fair verdict in a court case, so also we may trust that a neutral account of a dispute will maximize the reader’s chances of believing something true.

What, then, is the analogy to biased text? I propose that it is the elimination of an adversarial jury trial altogether. In the inquisitorial system, the judge leads the trial, never engages a jury, and determines the verdict from the bench, speaking from authority. And this is essentially what biased texts do: the ultimate judging role of the individual reader is dismissed. How justice plays out depends on the abilities of the judge, but generally, the details of a position are not covered if the judge deems that position implausible. Perhaps justice, or the truth, will be arrived at. However that might be, the judge, or author, decides for us all, thereby infantilizing us and failing to respect our autonomy and dignity as free people.

12. The evil of totalitarian propaganda as takes the badness of bias to an extreme

There is yet another way into the same argument.

We agree that totalitarian propaganda is wrong. We resent totalitarian personality cults and indoctrination in schools and mass media. The Nazi propaganda minister, Joseph Goebbels, is generally thought to have been among the worst of the worst Nazis. There is surely something deeply alarming and oppressive about being expected not only to say certain things but also to believe a certain way, on pain of punishment by the authorities.

We in the liberal West regard these episodes with horror. Their horror is not just a reflection of the horror of mass killing and the threat of concentration camps. There is something soul-killing about the propaganda, the cults, the indoctrination itself.

I think the Goebbelses of the world, the puppetmasters who perpetrate totalitarian propaganda, have tremendous contempt for our right to make up our minds for ourselves. The propagandists do not want us to think for ourselves; they want to do our thinking for us. An independent mind—an enlightened one, in Kant’s sense—threatens the regime.

Totalitarian propaganda seems obviously bad; but why is it?

The answer, my arguments suggest, is that propaganda offends against human dignity because it does not respect our intellectual autonomy. If it is an essential part of human nature to be able to deliberate and reach our own conclusions, if that gives us our freedom, our dignity as human beings, then totalitarian propaganda is a deliberate and systematic attempt to totally deny us our dignity and our full humanity.

But if this is correct, then the error of more ordinary bias—journalistic, encyclopedic, and pedagogical—is no different, except in degree, from the crime of totalitarian propaganda. Nazi propaganda strikes horror in us because, in treating us as mindless drones, its offends against human dignity. If so, shouldn’t we be horrified, albeit to a smaller degree, when a journalist comes out in favor of a certain proposal, repeating only one side’s most persuasive talking points and failing to interview anyone on the other side?

13. Neutrality’s democratic consequences

There is one last variant on the theme of autonomy: it is that neutrality fosters democracy, a system in which power is ultimately vested in individual citizens. The consequences of neutrality are not just epistemological and moral, but also social and political.

Journalists have the franchise like everyone, and naturally they individually favor one party or another. Fully and fairly stating the evidence and arguments of all political parties, not just their own, doubtless harms their own party’s chances in an election. Still, the practice of neutrality is very important, for it has deeper and more important consequences than the temporary victory of this party or that.

I want to maintain that the robust practice of neutrality by the fourth estate strengthens democracy itself. It keeps power more firmly in the hands of independent, individual citizens, and out of the hands of an elite. It does so in remarkably many and diverse ways.

Neutrality requires that political journalists give a fair and colorful account of the drama of policy debates. This practice informs voters, allowing them to make fairer choices. It makes voters more anxious, as they should be, about the necessity of actually being informed when they do vote. It calls attention to and raises interest in substantive policy debate, causing more of such debate to occur. It calls attention to the details of policy and the evidence for and against positions and thus, one hopes, improves everyone’s quality of reasoning. It causes voters to reflect on that substance and not just react to slogans and personalities. It raises expectations that politicians ought to be serious thinkers and not just empty suits. It causes us to take our leaders and our fellow voters more seriously.

Last but certainly not least, it makes the practice of politics more intellectually complex, making us better appreciate the sort of critical, liberal arts education that is necessary to a robust democracy. It gives us an added incentive to mold our young citizens into well-prepared participants at all levels of the process, which can only be to the good of everyone.

All of this is contrary to the ethos of an authoritarian society. Journalists, encyclopedists, and teachers all bear a very heavy responsibility to fight against the authoritarian tendency—by being scrupulously neutral as they report the debate and not just the horse race.

14. The argument from the Golden Rule

My second argument is much simpler: the Golden Rule holds that we should treat others as we want them to treat us. (For philosophers, I might just as well put this in terms of Kant’s First Formulation of the Categorical Imperative, for example. Interestingly, perhaps, my first argument may be cast in terms of Kant’s Second Formulation.)

Suppose you’re writing an encyclopedia article about some topic on which you have a strong opinion, such as politics, religion, or sports, and you have to represent an opinion you think is completely wrong.

Well, the Golden Rule asks, how would you want to be treated in this situation? So let us imagine that someone else is doing writing the encyclopedia article, and you are the reader. The issue is one you are fanatical about—but the writer takes the other side. Perhaps you’d like them to ignore their own opinions and repeat yours. But that isn’t going to happen: they disagree with you. You wouldn’t want to put aside your own opinion and repeat theirs, either. At the very least, you’d like your opinion represented in this encyclopedia article, not ignored. Indeed, you want it explained fully and fairly. In fact, you want the best arguments and the strongest evidence offered. But you’ll grudgingly admit that other opinions have to be given the same treatment, as long as yours is as well.

So the Golden Rule says that that’s how you, too, should lay out opinions you disagree with. As a result, you’ll paint the whole dialectical landscape in its most vibrant colors, and not just one part of it.

In short, when people set themselves up to be authorities, condescendingly telling you The Truth when you might want to disagree, you naturally find it irritating. You want to be treated like an adult, presented with all the best arguments, so you can make up your own mind. So, since that’s how you want to be treated, the Golden Rule would have you treat your readers that way, when you sit down to write a traditionally neutral text.

15. The argument from cooperation

So far, I have given two main arguments for neutrality, which are conceptually related: first from autonomy and second from the Golden Rule. A third, from cooperation, and fourth, from reliability, I can state much more briefly.

Here’s an argument from cooperation.

Wikipedia’s predecessor, Nupedia, originally adopted a neutrality policy, which Wikipedia inherited, partly because neutrality provides a way to defuse potential problems when people collaborate. If people are working together on the same text, the text can’t take a controversial position without potentially alienating some co-writers. Of course, if the collaborators share the same views, this is not a problem. But if the collaboration is, as on Wikipedia, open-ended—if you don’t know who, in the future, will be working on it—then the prudent way to avoid conflict is simply not to let the text take a position on controversial issues. Insisting that shared text in a collaborative project remain neutral provides the basic diplomatic framework that enables such projects to exist without permanent ideological warfare.

But since this argument is relevant only to open wikis and other collaborative works, that’s all I’ll say about that here.

16. The argument from reliability

Let’s consider one last argument, from reliability. The basic claim is simple: neutral writing is truth-conducive. Neutrality makes it more probable that readers will arrive at true and nuanced beliefs. (Compare the following argument to Chapter 2 of J.S. Mill’s On Liberty.)

The conclusion here is that the exercise of rationality is truth-conducive: we are more likely to have true beliefs if we base those beliefs on cogent arguments and solid evidence. Neutral writing supports the exercise of rationality by presenting us with, ideally, a full battery of arguments, with all the relevant evidence on all sides. This better equips us to determine what is true than if we receive only one side ex cathedra.

We’ll be more likely to have true beliefs if, as a matter of policy, we are presented with neutral writing. The more points of view we consider, and the more fairly and sympathetically the arguments for the various competing views are represented (even if they are wrong), the more likely we will be to arrive at true beliefs as a result.

17. A shotgun has a better chance to hit a fuzzy target

Some people might find this to be wrong or puzzling, in cases in which the truth is known. Ideally, they’ll say, of course the best way to ensure that a reader has the truth is to determine in advance what the truth is and then present it and only it in all its glorious detail. Generally, I find nothing wrong with that, as far as it goes; however, it doesn’t go very far, as it assumes the truth to be known and—this being the sticking point—uncontroversial. What if there is significant disagreement about what the truth is, and undecided readers know this?

In such cases, undecided readers have no rational grounds on which to prefer one source over another, everything else being equal; if they wish to follow Kant’s advice, sapere aude, they must make up their own minds. If they see different experts and credible resources saying different things, which are they to believe?

What if there are five different expert views on a question, each having adherence of 20% of the expert population and a similar chance of being right? Suppose an article advocates only one of these, and it turns out to be the wrong one. Then the chance of the article’s readers reaching a true belief, based on this biased account, is 0%. But if we canvass all five views, at least those readers who settle on the correct view will have a true belief.

More generally, if one side is ignored, or made to look very bad, and that side turns out to be correct, readers are discouraged from learning it. That is what my young friends would call a “fail.” And even if the truth happens to be presented with a favorable bias, it won’t be presented in its strongest, most rationally persuasive form if the reader can’t compare it to the other, wrong side.

Suppose instead that I have indeed made up my mind about something and, fortunately, I am correct; still, neutral sources inform me about other, incorrect views. That often helps me to understand and support my own position better.

In short, a neutral presentation at least exposes readers to the truth (if anybody has it) and, by comparing it with false views, enables them to understand it better.

18. A complete map of the dialectical landscape

It’s more complicated than that, though, of course.

Frequently, there are not just two bare claims on some narrow question; instead, there are various competing webs of interdependent claims, which are best weighed at the same time. The claims, taken together, make up what I might call the “dialectical landscape” surrounding a position: all the arguments, facts, details, attacks on the other side, rebuttals to their attacks, history, and context of some core proposition. For example, the dialectical landscape of a typical Democratic Party approach to the Affordable Care Act is very different from that of the Republican Party. Whatever you think of them, both are complex, both involve many subtopics and specialized claims, and they don’t even always address the same questions.

Truly neutral text, then, fully maps out the dialectical landscape on both, or all, sides of an issue, in a way that the sides can recognize as fairly representing their views.

A map of the opposing camps provides an invaluable service to us undecided interlopers. This isn’t just because we have a better chance of deciding rationally whether the core claims are correct, but also because the quality of our understanding is greatly improved the more familiar we are with all parts of the landscape. Even if we finally embrace some false claims, our minds are improved by being exposed to the context of all the claims. After all, in the dialectical landscape surrounding all competing claims are typically many others that are uncontroversial and yet very important to evaluating both sides.

Biased writing, by contrast, tends to be comparatively simple, too often appealing to emotion and employing fallacies and simplistic versions of arguments. Even if a piece fully and fairly canvasses one side, a biased account that mentions the other side makes a caricature of it, often simply to make its own side look more reasonable. Such caricatures propagate falsehoods—they are false portrayals of what the best representatives of the other side actually believe. But the point is that such caricatures actually stand in the way of improving the sophistication of our acceptance of our own side.

Arguments against neutrality

19. Examples of “good enough” neutrality exist

I want to deal with some objections. There are two big ones: (1) neutrality is impossible, and (2) in some cases, it is simply the wrong approach.

When defending neutrality as a policy, I often hear the following. We all have our views; nobody is unbiased. Those views are bound to come out, one way or another, in any exposition of a controversy. We can try to be unbiased, but we will always fail. It’s inevitable.

Yet many of us have had the experience of taking a class from a teacher who presented both sides of controversial questions sympathetically, and at the end, the teacher asks, “So, what do you think my view is?” And nobody could tell. I taught a philosophy of religion course once and at the end of the course asked for a show of hands: “Who thinks I believe in God?” About a third of the class put their hands up. “Who thinks I believe God does not exist?” About the same number of hands went up. “Who thinks I am agnostic?” Again, about the same number of hands. That was a proud moment for me.

Similarly, many of us have—I certainly have—had the experience of reading a news or encyclopedia article covering some debate, and it occurs to us that we cannot tell what view the writer endorses. The article seems admirably even-handed.

Perhaps it is true that one can always, in any text above a certain length, find something to improve as regards neutrality. Neutrality is extremely difficult to achieve—a point I will elaborate toward the end of the paper. But “good enough” neutrality, neutrality robust enough to earn our praise, plainly exists because we have seen many examples of it.

There is a different way into the latter objection. One of the main insights that motivate some people to say that neutrality is impossible is that objectivity is impossible. We all have a point of view, goes the objection, and this makes claims that certain propositions are “objectively true,” or true independently of any observer, very problematic—perhaps indefensible. (This issue is discussed in Thomas Nagel’s well-known book The View from Nowhere.)

I doubt that this is a problem for claims about neutrality, however. As I said above, “objectivity” and “neutrality” are two quite distinct concepts. The first is a quality of beliefs or attitudes; the second is a method of or constraint on exposition. I can be as subjective and “biased” as you like, in my heart of hearts, even as I am crafting a supremely neutral text. Similarly, the claims made in such a text can be neutral, fair, balanced, or untendentious without being objectively or subjectively true or false. The point of neutrality is to be fair to what people actually believe. One can be fair despite taking a side.

20. Chomskyan cynicism

Some people claim that neutrality is impossible basically because they have a cynical view of the world. To them, neutrality is always a front, a show—never sincere. Politics, like life, is nothing but a power struggle, they believe, and any claims to neutrality are instead just a cynical device for smuggling an ideology into a reader’s thought-world under cover of such precious epithets as “fairness,” “objectivity,” or “neutrality.”

Given the abundance of examples of apparently neutral writing, such a cynical stance seems unwarranted and merely puzzling. How can these people justify their cynical attitude in the face of such seemingly obvious examples to the contrary? What are they thinking of? They seem too cynical. Perhaps this is how they think; but it is most assuredly not how everyone thinks.

I find in conversation that the cynics frequently defend their views in terms of Noam Chomsky’s influential book Manufacturing Consent. Chomsky and his followers state that the news media may be given a “propaganda model,” i.e., the news serves merely propagandistic purposes for our corporate masters. They argue that the various “filters” that make up this model put the lie to pretensions of journalistic neutrality. Basically, the news is filtered: some legitimate stories are excluded, and others are distorted by mechanisms of “money and power,” including such things as the size of mass-media firms, advertising, reliance on establishment sources, “flak,” and “anti-communism.”

The most common thing one hears, along these lines, is that the big news media organizations are owned by corporations and thus inevitably biased in favor of corporate interests—for example, ABC is owned by the Walt Disney Company, Comcast owns NBC, and National Amusements is majority shareholder of CBS. If corporate interests are in control of the news media, how could the news fail to be biased in favor of such interests and of capitalism generally?

21. Chomskyan cynicism is unconfirmed and doesn’t establish the thesis

There is no brief way to address Chomsky’s argument head-on, so I won’t try. Instead, I’ll simply discuss this type of argument. The difficulty with it in this context is that it is a hypothesis formed more or less a priori—the news must be “filtered” because it is subject to the mechanisms of money and power—and while the hypothesis might be true (although, frankly, I doubt it), it can be confirmed only empirically.

What would go into confirming the claim? At a minimum, one would have to find some way to identify instances of bias and then confirm that that bias tends to be significantly associated with Chomsky’s “filters” such as corporate interests. The question then devolves from the rarified, abstract air of “systemic bias” to the more humdrum, concrete instances of bias that can be detected in actual reporting. Even if Chomsky and his followers were able to marshal some convincing examples of bias in corporate-sponsored reporting, the fact remains that there are plenty of examples of unbiased reporting. At least, in my opinion, there have been lots of news articles that lack any significant amount of bias.

Even if it is completely correct—and falsifiable—Chomsky’s cynical hypothesis appears to be that there is one source of thoroughgoing bias due to the influence of “filters.” It does not follow that neutrality is in principle impossible. In other words, even if we can identify something like “systemic” bias, or the bias of the whole (unreconstructed capitalist) system, it does not follow that there is bias in every instance of every publication, every article, every paragraph, and every sentence.

Another response is simpler: Chomsky’s criticisms concern only corporate news. What about private, independent media, individual and small bloggers, academic and independent encyclopedias, and various non-corporate textbooks? Even supposing that corporate news is hopelessly and systemically biased, that would not establish that neutrality per se is unachievable.

22. Is neutrality relative?

Next I will address a different sort of argument that neutrality is impossible. Its basic premise is that neutrality is relative.

This is hard to dispute: it seems to follow from the very definition of neutrality. If neutrality is a matter of fairly representing the various opinions on a question, one is right to ask which opinions should be included. We routinely exclude the extreme minority, idiosyncratic, and personal opinions from traditionally neutral texts, and we don’t call them biased for this reason. But we also routinely exclude (or fail to weight) opinions from foreign cultures. And for this reason—I will have to agree—our texts are typically biased in favor of our own cultures.

A New York Times article about the American policy debate over gun rights and gun control, which is remarkably even-handed in its treatment of Republican and Democratic approaches, correctly looks extremely biased to a Briton, for whom the very idea of “gun rights” seems ridiculous and horrible. And so it is. The Times article is biased: it carries an American bias. One may also say that an encyclopedia article for the Catholic Encyclopedia might be unbiased with respect to liberation theologians and conservatives, while remaining biased, as one would expect, in favor of Catholicism generally. And the same may be said for textbooks and encyclopedias and so forth: even if they are neutral with respect to a particular community, they are biased in favor of the more distinctive aspects of the community’s outlook.

This is interesting, but it is not very obviously a serious objection to a neutrality policy. It might cause us to review the arguments in favor of the policy. Indeed it seems we are indoctrinated, by our newspapers, encyclopedias, and textbooks, into our community’s basic assumptions. Are we less free for that indoctrination? Cosmopolitan philosophers like Descartes (Discourse on Method, Part 1) and Spinoza (his Ethics again) certainly thought so—to say nothing of contemporary philosophers like Martha Nussbaum and Kwame Anthony Appiah. I think so too. But it is surely too much to ask that writers somehow take account of the enormous range of views from across the whole globe when they write in traditionally neutral forms—or even always to care about those views. Arguably, it is human to be part of a culture, and, globalism notwithstanding, there is really no such thing as a “global monoculture” with respect to which writers should be expected to be neutral. Maybe that will exist someday, but it doesn’t yet.

I suppose a committed multiculturalist or relativist might want to take me to task for this, saying that it would surely be a grand thing that we be exposed to the broadest range of views from across the globe and across all ideologies, religions, and philosophies. (But what about eras? Shouldn’t we canvass historical views as well? We are surely present-biased.) Such a perspective is what one hopes to gain from international travel and a broadening, liberal education. But with the possible exception of remarkable polymaths, probably none of us has such a global perspective that we could make sense of a globally “neutral” newspaper, encyclopedia, or textbook—one that comprehensively takes account of an entirely global range of views.

Wikipedia itself might be thought to be committed to such a completely international neutrality, and in places, its policies have seemed to hold it to that utopian ambition. But of course it cannot be and it is not. The English Wikipedia’s articles about science most clearly betray its Western and especially Anglo-American provenance, and articles about, for example, philosophy are mostly about Western philosophy. I see nothing really wrong with that. There are many pages that report comparative information about conditions all over the world; but I have never seen a page that actually tries to synthesize and compare opinions from around the world, about controversial questions. It might, perhaps, be a worthy goal to create a new version of Wikipedia that is fully committed to being internationally neutral. If it succeeded, it would be the world’s first and only encyclopedia (or traditionally neutral text of any sort) that is fully neutral. I’m not claiming that many people would find it to be very useful. Maybe it would, but I might instead prove to be a curiosity of interest only to a few scholars.

23. Are some positions unworthy of inclusion in a traditionally neutral text?

The last major objection to a neutrality policy that I will consider is perhaps the most potent for many journalists: sometimes, certain points of view simply don’t deserve inclusion in a traditionally neutral text. Surely, some opinions are simply wrong—scientifically disproven, grossly offensive, or just plain idiotic.

Journalists sometimes make this objection. They react with exasperation at accusations of “liberal bias” from conservatives, arguing that sometimes, since one side of a dispute is clearly false, any balance would be “false balance.” The mission of journalism is to uncover the truth, and its first commitment is to accuracy. Accuracy, these journalists say, forbids us to strike a spurious balance between what are in fact two very unequal views.

I think some journalists can use this device as a cover for naked partisan bias: if there is a legitimate debate and an activist journalist refuses to find a “false balance,” that’s a fallacy. Whenever journalists choose a side, on grounds of avoiding “false balance,” they owe us an argument that the disfavored side really does not deserve neutral treatment. What sort of argument would do the trick?

Consider some points of view that mainstream journalists often find unworthy of uncritical coverage: global warming skepticism; creationism, intelligent design, or creation science; the anti-vaccination movement; homeopathic medicine. Those examples are from science, and similarly controversial examples might be drawn from recent events and history: Holocaust denial; “birtherism”; the 9/11 conspiracy theory; the Bilderberger or Jewish banker conspiracy theory.

The popularity of such views combined with the relative silence about them in the media suggests that journalists would find neutral coverage of such topics to be “false balance.” These also suggest at least two different variants of the argument to consider: a moral variant and an epistemological variant.

On the moral variant, the reason we should not fairly represent certain positions is that the positions are morally reprehensible, harmful, and/or dangerous. This is a common German attitude toward Holocaust denial. Moral concerns, therefore, outweigh society’s interest in neutral reporting.

On the epistemological variant, the reason is that the positions are so unscientific or otherwise ill-supported that it simply insults readers’ intelligence for a writer to treat them seriously. This is a common scientific attitude toward homeopathy and global warming skepticism. Here, epistemological concerns outweigh society’s interest in neutral reporting.

In either case, the argument is simple: certain views are either morally or epistemologically bankrupt to an extreme degree. So, the claim goes, traditionally neutral texts are under no obligation to include them. Therefore, since there are exceptions, neutrality is not a universal, absolute principle. It has its place, my journalist friends tell me, but it operates within ethical and epistemological constraints—which they determine, of course.

24. The relevance of morally repugnant views depends on the context

This all seems reasonable. Heroically, I shall proceed to disagree.

Let me begin by conceding that there do seem to be instances of positions that I would want to exclude from an encyclopedia, for example. It it does seem I would want to do so on grounds that they are thoroughly (morally) reprehensible or utterly (epistemologically) ridiculous. Is shoplifting to be represented as a legitimate hobby? Do the attitudes of child molesters or sociopathic killers toward their crimes merit serious discussion? Do we really need to spend any time on Holocaust denial? For some, these seem to be rhetorical questions; but I won’t treat them that way.

In a list of hobbies covered by an encyclopedia, we wouldn’t include shoplifting, even though it serves as a hobby for some people. But there are perspectives in psychology or sociology that consider shoplifting that way. And while a general encyclopedia might not include shoplifting among the hobbies, probably we can justify that exclusion not on grounds that it is immoral or criminal but because it simply is not ordinarily classified among the hobbies.

Similarly, in news reports of crime statistics, crime’s impact on communities, and police and legislative strategies to defeat it, discussion of the attitudes of child molesters and sociopathic killers is an unnecessary and unwelcome intrusion. That is because they are irrelevant to the concerns of news readers. But criminal attitudes are certainly a focus of sociologists and criminal psychologists, who need to understand the motives of criminals; in criminology textbooks, criminal attitudes are crucial to include in chapters about criminal psychology. And sometimes, criminal motives are newsworthy. For example, we all wonder what impels a handful of inner-city blacks to shoot white police officers without provocation—and vice-versa.

Even Holocaust denial is open to the same sort of context-based analysis. There is no need to include Holocaust denial as an “alternative theory” about German history in WWII, for the simple reason that it is an extreme minority position; textbooks are under no obligation to include extreme minority positions. Of course it is repugnant, but its repugnance is not why it is not covered in textbooks. After all, there are circumstances—e.g., the politics of modern Germany, or German sociology—in which the views of Holocaust deniers, and getting the facts about them correctly and even sympathetically stated, might be important.

This discussion should make it plausible that it is not the moral bankruptcy of certain views that would lead us to exclude them from traditionally neutral texts, after all, but just considerations of relevance. It depends on the purpose of the texts. In some sorts of texts, thoroughly reprehensible things are described in stunning detail. In others, they simply aren’t relevant.

25. The same sort of analysis given to teaching creationism in schools

So much for the moral argument. What about the epistemological version? Is such an analysis possible for those cases in which a position seems legitimate to exclude because (on the view of many scientists or scholars) it lacks any sound intellectual basis?

I can’t sidestep the issue when it comes to, for example, creation science or creationism in biology textbooks. I won’t pretend to be able to adjudicate this politically charged issue in a short space, but I will make a few germane remarks and draw some limited, tentative conclusions.

On the one hand, the public and their representatives in Texas certainly have the right to ensure that public school children are exposed to views about what fundamentalist Christians believe about the creation of the world and the origin of species. In the U.S., they do not have the legal right to teach only those views in public schools, because that would run afoul of First Amendment prohibitions of the state establishment of religion. But it seems perfectly unobjectionable that those views, if placed among others, are canvassed in public schools.

On the other hand, scientists have an excellent point when they maintain that creation science is an extreme minority view among scientists. Even scientists with a Christian point of view generally endorse evolutionary theory, as does the Catholic Church. If the purpose of teaching biology is to impart facts and theories that biologists stand behind, then biology classes should do so.

A policy of neutrality would have us determine what the relevant community is. In science class, that would seem to be the scientific community. But science class is taking place in the larger context of enculturation by the school system, and there the relevant community is the general public, not just scientists.

Bearing this in mind, I think adopting a policy of neutrality would entail the following recommendations:

1. Insofar as biology class is canvassing the views of actual biologists, creation science has no place there; for better or worse, the vast majority of scientists simply do not consider it to be a scientific theory at all, but a theological one. So it is simply incorrect to represent it as one view among many.
2. Insofar as biology class takes place in the context of a broader schooling context, however, the people of Texas (and other such places) have the right to insist, not that the science be taught a certain way, but that students be informed that a large number of citizens disagree with the science. Little time need be spent in informing them of this disagreement in science class, because the disagreement is not a topic in science but in the broader culture. But the dignity of students and their parents is best respected when their disagreement is officially and respectfully acknowledged.
3. In the interests of neutrality, in addition to evolution as taught in biology, students should be taught about (neutrally, of course) a variety of religious, philosophical, and scientific views about cosmology, creation stories, and other topics that religion, philosophy, and science all speak on. Such an interdisciplinary class or unit would be the place in which creationism would be canvassed. Needless to say, perhaps, students in that class whose parents do not want a neutral approach to this subject should not be forced to endure it.

This is a compromise solution, and I hope a neutral one as well.

I could give similar treatment to other issues. What I want to maintain is that for all the issues listed earlier, there exist sophisticated, often multi-part neutral solutions. In some cases I might seem to favor one part or another; but taken all together, the result will be fair. Journalists, or the state, will not unduly favor one position over another, affording readers and students the dignity and tools to decide important, controversial issues for themselves.

26. Neutrality and global climate change

A few comments about how to write neutrally about climate change should clarify some more issues.

A majority of climate scientists believe in anthropogenic global warming (AGW); exactly which scientists to include as “experts,” and thus which surveys to consider and what percentage endorse it, is a matter of debate. In any case, while there are some distinguished climate science experts who are skeptics, a decided majority are not.

When explaining the science itself—so, in science textbooks and science research news—understanding the controversy is not the primary aim. Understanding the state of the science is. So, presumably, in that context, most of the issues and most of the evidence will appear to be on the AGW side, not because the writers are biased but because most of the actual science happens to be on that side. In other words, if the mission of the text is to canvass the science, the text will appear biased but it will not really be biased; any bias is to be found in the subject matter itself. Besides, there are many skeptical qualifications and objections that will be worth covering, even if they are dismissed by most experts. One justification of covering them, apart from the requirements of neutrality, is that it helps the mastery of any very complex subject (like climate science) to consider objections and qualifications.

But matters are different in a political context—political journalism, or encyclopedia articles or textbook chapters about the politics of global warming. There, as important as understanding the science is, understanding the controversy is even more important. Coverage of such a controversy cannot be done without apportioning space at least roughly equally.

Climate change scientists and activists have dismissed the idea that journalists ought to report neutrally on climate change skepticism by arguing, “If you’re going to give equal space to climate change skepticism, you should also give equal space to Holocaust denial—which is absurd.” But I hope it’s clear how I would respond to that argument: while among the American population, climate change skepticism appears to be a (large) minority position, Holocaust denial is a tiny fringe phenomenon. In an American context, reporting neutrally about the climate change debate does not entail that reporters must so much as mention Holocaust denial. In certain other countries, however, things might be different.

I realize this section’s discussion is just a lot of hand-waving on my part, and more needs to be said. Beyond this, the issues are to be taken up by the experts themselves (those who really care about neutrality). That, as we will see next, is the point: the specific requirements of neutrality in any particular case can be discovered only by careful study of the details of the case. The devil of bias and the angel of neutrality are in the details.

The difficulty of neutrality and how we might eradicate bias

27. Neutrality is complex and difficult

After this long discussion, I hope it will be abundantly clear that neutrality is not a simple concept. “Why can’t journalists just be more neutral?” readers ask, complaining about “media bias.” But this assumes that neutrality is something simple and straightforward to fix, like a spelling mistake. It might appear simple, if you think of neutrality as no more than giving an equal amount of space to each position and avoiding obviously controversial judgments.

Apportioning space evenly and avoiding words that imply judgment is a good place to begin, especially when you’re reporting about a policy debate. But if your purpose is to construct a narrative about a multi-layered, developing situation, neutrality is instead a matter of relaying, while reporting the basic facts, that there are alternative views, mentioning them in the right places, providing different possible explanations, and so forth.

The failure to understand or acknowledge the complexity of neutrality is widespread, even among writers. That failure in turn leads many writers, even of traditionally neutral texts, to underestimate the challenges and fail to take practical measures to meet them head-on. Writers often fail at this and end up writing biased stuff without realizing it, even when they sincerely believe and intend otherwise. The sheer difficulty of neutrality is at the root of this failure.

When we learn to spot instances of bias and look hard for it, we can see it everywhere. For some, this in turn causes cynicism about the very possibility of neutrality.

28. If bias is the norm, it must be eradicated deliberately

Cynicism is understandable. We should expect bias to be the norm, if it isn’t deliberately eradicated.

That bias is the natural state of affairs became obvious to me after I attempted to herd the many biased cats of Wikipedia and Citizendium, and later compared those experiences to what I learned about how the admirably even-handed ProCon.org operates. (Disclosure: ProCon’s publisher is an investor in a startup of mine; but we got to know each other some years before that due to the very issue of neutrality.) ProCon’s executive team told me they have detailed rules about neutrality as well as a thorough process in which people from a variety of ideological perspectives give brutally honest feedback on the fairness of drafts. The result is remarkably even-handed articles.

Well, if ProCon has to invest so much time and attention to make their pages neutral, what makes the rest of us think we can do it without spending a similar amount of time and attention?

29. How publishers can improve neutrality

Publishers, I propose, need to add two features to combat bias: first, detailed neutrality guidelines, and second, specific procedures for making writing more neutral, including specially assigned people.

We would need a complex set of neutrality guidelines to identify neutrality problems and adjudicate issues. I haven’t included such rules in this document. To a great extent, the rules will resemble a style guide. Wikipedia has an extensive guide, and I have developed a guide for Ballotpedia. Other publishing organizations that claim to be neutral or fair ought to be able to articulate exactly what guidelines they follow in identifying and resolving neutrality problems.

Beyond simply having rules, staff need to follow some procedure to ensure that the publication as a whole, as well as individual articles and their components, are all neutral. Some publishers invest money into copyediting and fact-checking, presumably on grounds that it makes their publications more readable and accurate. Similarly, I propose that publishers of traditionally neutral texts ought to care much more than they apparently do about neutrality, and they might wish to assign editors and writers to handle the most difficult neutrality issues and otherwise implement their neutrality methodology.

ProCon goes further, constantly worrying about the issue and following arduous procedures to achieve it. Ballotpedia, too, has hired two people for a new sort of position that will help with neutrality issues.

30. The discipline of neutrality

Let me elaborate the second point here.

I am proposing that publishers add what would for many be a new layer of oversight: not just vague expressions of commitment to neutrality, but specially assigned neutrality editors, or some other way to more actively and aggressively edit for neutrality, in addition to the traditional disciplines of copyediting and fact-checking.

Ballotpedia recently hired its first writers specifically tasked to work on neutrality. I helped hire them. We had the top candidates evaluate the neutrality of a report Ballotpedia had recently done. I immodestly prided myself on my ability to spot problems with neutrality, and the Ballotpedia piece struck me (on a too-casual reading) as having only a few problems—nothing too bad. But after the candidates got through with it, it was clear that the report had many more neutrality issues than I had noticed. The problems were mostly subtle and perhaps understandable, but the criticisms all made sense. And more to the point, they were all fixable. If the writer of that study had had feedback from a neutrality editor, the piece would have been substantially more neutral and better-written as a result.

Similar instances of surprising, “hidden” bias are rife in most news reporting, textbooks, and encyclopedia articles. If we choose to view neutrality as a discipline similar to good spelling, then the bias of today’s journalism, textbooks, and encyclopedias looks clumsy and backward, like the weird spellings one finds in books published before spelling became standardized.

Professional publishing operations already take the time for the traditional disciplines of copyediting and fact-checking, because publishers and readers value readability and accuracy. For traditionally neutral forms, shouldn’t they also value neutrality just as much? The moral and political arguments in favor of a new discipline are as strong. If you do not take the time to edit for neutrality very explicitly, you will simply fail to be neutral. How could you not? The same can be said of copyediting and fact-checking: if you don’t take the time to edit explicitly for correct mechanics and accuracy, you’ll probably screw up. To err is human. If you try hard and are conscientious, you might do a decent job—maybe. If you want to be sure, hire some professionals and set up a neutrality editing process.

31. No newspaper has ever been neutral

If I am right that bias is rife and yet we have good reason to value neutrality, there is a great latent demand for reporting that is neutral in the sense described here. I know that I would take great interest in any news, encyclopedia, or textbook publisher that credibly announced a claim to neutrality in the sense I have defined.

The emphasis here is on “credibly.” Obviously, claims to being “unbiased” or “objective” have long been a cheap but meaningless selling point for various clearly biased publications. Idle claims to the effect of “we are already neutral, of course” are simply not credible.

After all, the reply is simple: No, you aren’t. No newspaper has ever been neutral. No newspaper has ever even tried properly, as a matter of policy.

Again, this isn’t a complaint about “liberal media bias.” Fox News claims to be “fair and balanced” but of course very few people believe they are. Traditional news broadcasts project a tone of objectivity and generally claim not to be biased. But nobody really believes they’re neutral in the sense defined in this paper. Again, I say this not because I’m accusing them of liberal bias but because neutrality is difficult and simply will not be achieved except deliberately, recognizing the perhaps surprising complexity and difficulty of the job and devoting the effort needed to get it done.

A publisher’s claim to neutrality could be substantiated by (a) publishing a statement of neutrality principles, (b) taking proactive measures to ensure that they are followed, (c) doing studies of reader and external neutrality auditors vouching for success, and finally (d) publishing the results of the audits.

32. There is an unmet latent demand for neutrality

There’s a great latent demand for neutral content, and the demand is unmet.

There will always be a market for biased reporting and opinion. But publishing operations that can credibly state that they embrace neutrality would have an unusual advantage over others. So I speculate that startups in a new “neutrality niche” might do well. There could be newspapers, weeklies, wire services, and more.

I can also imagine startups in reference and textbook publishing that could credibly make similar claims. As an encyclopedia aficionado and former college teacher, I know I’d be extremely interested.

Finally, I can imagine firms offering external audits of the neutrality of publications and research organizations.

Readers can help make this happen, too, simply by making their preference for neutral texts known. You simply need to promote the proposition that publishers should adopt both detailed neutrality guidelines and procedures for ensuring the guidelines are followed.[1]

1. For discussion and other help, thanks to Kamy Akhavan, Courtney Collins, Angela Consani, Anthony DiPierro, Scott Duryea, Stephen Ewen, Mike Forsythe, Leslie Graves, Sara Key, Greg Lukianoff, Geoff Pallay, Terry Phillips, Jay Rakow, Gerry Sanger, Kristen Smith, Jason Swadley, Anton Sweeney, and Bryan White.