User:Rosguill/New pages patrol is racist

I need to start off this essay with a number of disclaimers. This essay is not intended as an indictment of new page patrolling as a content control process on Wikipedia. As of this writing, I have been one of the most active new page reviewers for over a year, and have no intention of stopping––I clearly do not think that new pages patrol is a Bad Thing. Nor is this essay an attack on the attitudes of new page reviewers, whether collectively or individually: while Wikipedia has many structural problems, a high-powered cabal of white supremacist editors is thankfully not one of them. Rather, this essay is intended to examine how a group of well-intentioned editors working to the best of their abilities can nevertheless end up imposing patterns of behavior that produce outcomes that are, in a word, racist.

The problem

The underlying problem at NPP is one that has been documented across Wikipedia: as a product of a world rife with systemic bias, Wikipedia itself reflects this bias in its content. NPP in particular, however, is an important checkpoint where a significant amount of this bias is directly enforced. The combination of the relative lack of availability of sources that write about marginalized groups and the high likelihood that new page reviewers will not have any particular expertise or insight into subjects important to marginalized groups that would let them assess notability beyond what is readily available in the article or online together mean that a disproportionate amount of articles about such topics will end up wrongfully deleted, and we are unlikely to even recognize that these deletions were wrong as they are occurring.

(This disparity is not just racism: in many cases it is better described as eurocentrism, anglocentrism, sexism or other forms of chauvinism, but titles need to be concise).

Part of the problem is that correcting for this bias isn't as simple as just being more lenient when reviewing articles about marginalized subjects: relaxing our notability guidelines at a whim could lead to the inclusion of misleading, promotional or otherwise harmful content, which is not an improvement over the absence of content. Addressing our systemic biases requires careful thought about what content is actually harmful to Wikipedia's mission, and an assessment of when we can afford to give articles the benefit of the doubt.

Working towards a partial solution

A proactive step for addressing this issue would be to relax notability guidelines for articles about subjects affected by Wikipedia's systemic bias that are unlikely to include content that is genuinely harmful to Wkipedia's purpose of providing verifiable information. In a sense, it's easier to identify the cases where we can't loosen our standards than places where we can.

In my opinion, the categories of articles where we have no business loosening our standards are:

1. Biographies of living people
2. Articles about contentious topics
3. Articles about unreleased media

I think that it's fairly self-evident why we can't relax our notability standards for biographies of living people and for contentious topics. It is important to recognize, however, how broad the category of contentious topics can be. While this obviously includes battleground subjects such as politics or ethnic groups, this can also include articles about food and other cultural practices (which can be a sticking point of ethnic or national identity). The decision to give the benefit of the doubt in articles about such subjects ultimately must be determined on a case-by-case basis. Finally, articles about unreleased media often have an element of promotionalism to them, in addition to having a particularly weak claim to notability by virtue of them not having been released yet.

For any subject that doesn't fall under the above categories, I think that we can consider giving an extended benefit of the doubt. This doesn't mean that we should instantly accept any article about a marginalized subject regardless of quality, but rather that we should consciously be more generous when assessing the likelihood of there being additional coverage in reliable sources that are not available to us. Rather than seeing this as a double standard, I think that it's best to understand this as an expansion of Wikipedia:NPOSSIBLE for subjects that we are ill-equipped to research. Consequently, it's worth noting that this additional leeway should only apply to articles about subjects that you yourself have little familiarity with. If you are intimately familiar with Bhojpuri cinema, there's no need to give more benefit of the doubt above and beyond your own assessment of whether significant coverage exists. For the rest of us that cannot read Bhojpuri and know little about the regional film industries of India, we should proceed with caution. Even one example of a professional review may be enough to suggest that sources exist for such a subject.

Finally, I think it's important that we be open about this with editors who are writing such articles, either by leaving talk page messages or using {{Notability}} tags. Otherwise, we run the risk of giving people a false impression of our verifiability standards, which could cause misunderstandings, conflict or just poor contributions further down the road.

Behavior at AfD

Another practice that I think would do us well to consider is that there is no obligation for us to be the most effective prosecutors that we can be once an article is already at AfD. While you should write a clear opening statement that identifies an article's shortcomings, it's important to remember that the goal of AfD is not to win. From the perspective of new page patrol, once you have decided to send an article to AfD, you've fulfilled your job: now it's the community's turn to decide whether the article should be kept or not. As a new page reviewer, you are likely going to be better at making AfD arguments than the editors who have contributed to the article, and thus will often be in a position where you could "win" the AfD even if a stronger keep argument could have been made than what was actually presented in the discussion.

There are nevertheless cases where an article has significant problems that warrant responding to keep votes, and it's also worthwhile to engage with additional sources provided over the course of the discussion. Obvious promotionalism, original research, and POV-pushing content forks should be challenged. But articles that have no significant flaw other than falling a bit short of GNG do not require relentless prosecution, and the underlying biases described elsewhere in this essay suggest that vigorously arguing for deletion at AfD is likely to produce results that further entrench bias.

Another problem

There is a second problem with racism at NPP, and that is our treatment of editors, and particularly new editors, who belong to marginalized groups. Anecdotally, editors with African or South Asian names are far more likely to be accused of COI or paid editing. Editors who do not have English as a first language also face discrimination, and sometimes end up with indefinite CIR blocks.

I don't have anything that I would consider to be even a partial solution for this problem. In the moment, editors requesting COI disclosures will be doing so because they genuinely believe that there may be COI at play; it is primarily our unconscious biases that make for an uneven playing field, and the unconscious nature of these biases makes them difficult to confront. I can only suggest that editors think hard about their decision to accuse someone of paid editing or other COI, and to consider leaving a short, personal note rather than a bureaucratic template, as the latter can be a considerably more alienating experience for the editor on the receiving end. Still, even if we can't draft a guideline that will fix this problem, I think that it's important for us to be aware that this problem exists, and it would be remiss for me to title an essay "NPP is racist" and omit this thorny aspect of our work just because I can't propose a solution.

Conclusion

While decisively eradicating racism and systemic bias is something that is out of our hands as Wikipedia editors, I think that the above practices, if adopted, would be a positive development for Wikipedia. New page reviewers are actively engaged in gatekeeping newly created articles. It's important that we recognize our own shortcomings when analyzing articles that are outside of our fields of experience, rather than imperiously deleting content that doesn't completely meet our guidelines, guidelines that, when assessed on a truly global scale, are structurally unfair.

See also

• Wikipedia:Systemic bias
• Wikipedia:WikiProject Countering systemic bias