Since the beginning of the year, among readers of HBD-related blogs there’s been an increased interest in trying to improve articles related to race and intelligence at Wikipedia. The appeal of this is easy to understand, since very few books or articles written for a popular audience accurately present the scientific consensus about this topic. (Some pop-science books about it, such as The Mismeasure of Man, are infamous for the amount of misrepresentation they involve.) On the other hand, while peer-reviewed journals and academic textbooks often do present this consensus, these generally aren’t accessible to lay people. Wikipedia is the most commonly-used source of information in the world, so if it can accurately represent how this topic is presented in the professional psychology literature, it’s providing a service that’s provided almost nowhere else.
However, there’s a problem. Many of these new editors don’t realize that contributing to any topic at Wikipedia requires more than just knowledge about the topic itself–it also requires an understanding of how to conduct oneself there. This is especially true in articles on controversial topics, and race and intelligence is the ninth most controversial topic at Wikipedia. Over the past few months, several of these editors have come to me for advice about how to make a positive difference in this area, so I’ve decided to write up some of my advice for them in a blog post here.
One point needs to stated before anything else: The most useful thing anyone can do on these articles is keep watch over them, and undo vandalism when it happens. The largest problem with most of the articles is that they aren’t being watched by anyone who cares about them. This means that people can remove chunks of sourced material without giving a reason, or add material without a source, and the changes often won’t be undone.
This problem didn’t always exist. Several years ago, there was a robust community of editors who maintained these articles, but many of them eventually quit out of frustration because of how they were being treated by one particular Wikipedia editor: Antony Wassermann, who’s better known by his Wikipedia handle Mathsci. For anyone who’s interested in delving deep into the history of Wikipedia conflicts, there’s a detailed history of this problem that I wrote here to assist one editor whom Mathsci was harassing in October 2013. (Note that this document requires a certain amount of familiarity with Wikipedia terminology.)
It’s easy to get a gloomy perception of what participating in Wikipedia is like from that document, but the situation has improved considerably in the time since I wrote that. The final instance of harassment described there, against an editor named Charles Ainsworth, resulted in Mathsci being banned from Wikipedia. During the time since his ban,
two (Update: three) of the editors who he directly or indirectly caused to leave the project have become active again. However, none of these editors have the interest and ability to participate in race and intelligence articles anymore, and all of the others have yet to return.
To anyone who wants to repopulate these articles, the following are some practices that it’s a good idea to engage in, as well as mistakes that should be avoided. Some of these are mistakes that I’ve seen other editors make, while others are mistakes that I made myself when I was a newbie there, before I knew any better.
Don’t give excessive weight to low-prominence sources
New editors often think it’s important to keep Wikipedia articles up-to-date with all of the latest sources, but this usually isn’t the right approach. The goal that Wikipedia strives for, with varying amounts of success, is to present each idea with about as much prominence as it has in the source material. (This is known as the neutral point of view policy.) With a few exceptions, newly-published ideas are not very prominent, because other researchers haven’t yet had the opportunity to build upon these ideas and incorporate them into their own research. Giving excessive weight to recently-published material is sometimes referred to as recentism.
Davide Piffer’s 2013 factor analysis paper is an example of recently-published research that some readers of this blog have suggested should be cited at Wikipedia, but that probably isn’t prominent enough. In the future this will likely be regarded as a historically important paper, because it’s the first study to directly show that the distribution of alleles linked to cognitive ability varies between ethnic groups, and does so in a way that matches the distribution of IQ scores. However, this paper was published in The Mankind Quarterly, which is a specialist journal with a fairly low impact factor, and according to Google scholar only three other papers are citing it. This means that by the standards of Wikipedia, the paper is not very prominent.
Do use mainsteam sources
In general, the best sources to use about race and intelligence are those attempting to impartially summarize the debate on this topic, written by people who aren’t active participants in the debate but who still are experts in the relevant fields. (These sorts of sources are called “secondary sources”, because they’re one step removed from the debate.) Two examples of these sorts of sources, which the articles already make considerable use of, are Earl Hunt’s 2011 book Human Intelligence, and Hunt & Carlson’s 2007 paper “Considerations Relating to the Study of Group Differences in Intelligence”. A source that’s somewhat older but still authoritative is Intelligence: Knowns and Unknowns, the official statement of the American Psychological Association.
It’s worth noting that when these sources discuss the cause of racial IQ gaps, all of them take a position that’s in between the 100% environmental position advanced by Nisbett and Flynn, and the 80% genetic / 20% environmental position advanced by Jensen and Rushton. For example, the position of the APA report was that there was too little data to identify the cause of the gaps, because there was a lack of direct support for both genetic and environmental causes. Earl Hunt’s book takes the perspective that Nisbett’s position is very unlikely to be true, but that the percentages given by Jensen and Rushton are too precise to be supported by data. In general, this intermediate perspective is the one most commonly taken in mainstream sources.
I’ve been asked at one point, “How can we get the truth into Wikipedia?” As I said in my response here, this is asking the wrong question. There are many ways that Wikipedia’s articles related to race and intelligence need to be improved, but holding them up against an objective standard of truth isn’t the best way to measure that. The way Wikipedia presents ideas is meant to be based on how they’re presented in the source material, so articles there can never be more accurate than the source material is; that’s a fundamental limitation of how Wikipedia works. As long as the professional psychology literature presents the hereditarian hypothesis as only one of a few possible explanations for racial IQ gaps, Wikipedia must present it the same way.
Don’t repeatedly make the same change that’s being undone
Suppose you make a change to an article that you think is an improvement, and someone else undoes it. What then? Inexperienced editors sometimes assume the answer is to make the same change a second time, and then a third time and a fourth, no matter how many times it’s undone. Doing this is a mistake known as edit warring. People who revert an article to an earlier version more than three times in a 24-hour period often are temporarily blocked as punishment; this is known as the three-revert rule.
As an example of what to avoid, consider DavidJac123’s edits to the article on Richard Lynn’s book Race differences in intelligence. He made the same change nine times in about two days, and as a result was blocked for 48 hours. (This article is being considered for deletion, but the list of DavidJac123’s edits is archived using WebCitation, so it should remain accessible even if the article is deleted.) Anyone who approaches this issue the way he did is likely to get a similar result.
Do use Wikipedia’s dispute resolution processes
When other people oppose your edits, the first thing to try is discussing the issue on the article’s talk page. Some issues are simple enough that this is all it takes to resolve them, although a lot of editors involved in these articles don’t change their minds easily. Still, you should always make an attempt to justify your perspective there. Wikipedia has a few other dispute resolution processes for when the talk page isn’t enough, but people are generally expected to only use them if they’ve tried to resolve it on the talk page first.
A request for comment is the most useful of these options, and every new editor on these articles should learn how to use it. When a person starts an RFC, their request is copied to a page soliciting input from other members of Wikipedia. Controversial topics are sometimes dominated by small groups of users with a vested interest in supporting a particular viewpoint, but a request for comment can attract outsiders who will have more impartial perspectives.
A second place that can sometimes be used to resolve disputes is the Dispute resolution noticeboard. The downside of this avenue is that it only works if there are approximately equal numbers of people on either side of the dispute–otherwise the side that has the majority can simply dominate the discussion, and cause it to become no different from the discussion on the talk page. A third option is to request mediation. The downside of mediation is that it can be quite time-consuming, and mediation requests usually are only accepted after there’s been a lengthy and unsuccessful attempt to resolve an issue on the talk page. You should only request mediation for important issues that you’re prepared to put a lot of time into.
The final option on any of these articles is arbitration enforcement. Arbitration enforcement is not a place to report disputes over the content of articles, and AE reports that appear to be content disputes usually are turned down. Matters should only be taken to arbitration enforcement when there is a persistent problem with someone’s behavior–for example, if a person only responds with belittling remarks and is unwilling to discuss the article itself. The person making an AE report will typically have their own behavior examined as well, and might be penalized themselves if their behavior is as bad as that of the person they’re reporting. Because of this, making AE reports can be very risky, and I don’t recommend doing so unless someone else’s behavior is clearly violating a Wikipedia policy.
Arbitration enforcement threads require a link to the arbitration remedy that’s being enforced. In most cases the remedy being enforced would be this one, authorizing discretionary sanctions on all articles related to race differences in psychological traits. Discretionary sanctions are a special type of permission administrators can be given to punish bad behavior, which they’re often given in articles about controversial topics.
Don’t try to immediately overhaul high-importance articles
Race and intelligence is the most highly-visible article related to the race and intelligence controversy at Wikipedia, so many new editors who care about this topic assume that article is a good place to start. However, this usually is a bad idea. Because of how visible this article is, any change to it that’s even slightly contentious will attract lots of attention from other editors. Due to the exodus a few years ago of most the editors who maintained this article, the remaining people who watch it also are weighted very heavily towards opposition to the hereditarian hypothesis. These editors sometimes allow changes to the article that increase its coverage of the hereditarian hypothesis if the changes are made by experienced and trusted editors, but never if the changes are made by newbies.
The one thing that is valuable on this article is to watch it and undo vandalism to it. Many Wikipedia editors only undo vandalism, or other obviously problematic edits, if the edits are opposed to their personal viewpoint. When that trend is combined with the current balance of editors who watch these articles, the result is that problematic edits that increase the articles’ coverage of the hereditarian hypothesis usually get undone very quickly, while similarly unhelpful edits opposing this hypothesis sometimes stay in the articles for months or years.
Fortunately, the main race and intelligence article isn’t one of the worse examples of an article where this has happened. While it’s an exaggeration to say that this article is in good condition, it’s not in nearly as bad condition as some of the sub-articles about other aspects of the topic. Since the problems on these sub-articles are both larger and easier to fix, new editors are far more likely to find that energy put into improving these articles is productive.
Attracting attention is something you’ll want to avoid for another reason as well. The more you attract the attention of editors who disagree with you, the more they will oppose any subsequent changes you try to make. When their attention is attracted to any part of an article, they also may make further changes to that portion of it, so that it eventually ends up worse off than it would have been if you hadn’t done anything. This scenario might seem far-fetched, but I’ve observed it happen numerous times.
Do start small
Most of Wikipedia’s articles related to race and intelligence can be found in the Race and intelligence controversy category, although there are a few articles related to it that the category doesn’t include. The following are five articles that need major work, but that have relatively few people paying attention to them. On most of these articles, the problems are the result of the trend I’ve described: problematic changes can accumulate indefinitely if the changes are favorable to editors’ viewpoints, because in that case they generally won’t be undone.
Race and crime in the United States – If I had to choose just one article related to race that needs major work, it’s this one. At the end of 2009, this article was in quite good condition, but in the time since then it’s gradually fallen apart. The ways in which this has happened are too numerous to list in this post, but I described them here at the Wikipedia criticism site Wikipediocracy. The article hasn’t had anyone maintaining it since 2010, so one thing that needs to be done is updating the article to include sources published in the past five years. A second thing that’s equally important is adding back some of the large chunks of content that were removed without any discussion.
Mental Chronometry – The main problem with this article is that it doesn’t discuss any of the research on Mental Chronometry that’s been conducted since the 1980s. The reason this problem exists is because most of the article was written by the user Ferahgo the Assassin in September 2010, in chronological order beginning with the early history of MC. In October 2010, before she had finished writing the article, Ferahgo received a topic ban prohibiting her from editing any articles related to race and intelligence. In the four years since then no one else has been motivated to finish writing it, so it’s remained indefinitely in this half-finished state.
Richard Lynn – I summarized the problems with this article in the third part of my post here. The reason the problems with this article are severe is because policies are supposed to be followed especially strictly in articles about living people, and these articles are covered by a special policy known as BLP policy. Fixing the problems with this article will probably be more difficult than in the first two articles, because there are a few editors who have tried hard to restore the policy violations when various people tried to fix them. However, BLP policy is unambiguous that in articles about living people, content policies cannot be compromised, and Wikipedia is not supposed to have any tolerance for editors unwilling to follow these polices in articles about living people. In addition to the dispute resolution options mentioned above, there is also a separate provision for behavioral problems on biographies of living people to be reported at arbitration enforcement.
Ashkenazi Jewish intelligence – The main problems with this article are its reliance on an unreliable source, and giving that source undue weight. Specifically, the problem is that one of the article’s two most-cited sources is this self-published paper by Brian Ferguson. According to Wikipedia’s verifiability policy, self-published sources are only acceptable when their authors are recognized experts, whose work in the relevant area has been reliably published elsewhere. However, Google scholar searches for publications by Ferguson containing the words either “Ashkenazi Jews” or “Intelligence quotient” both return no results except that single paper. Since he’s never written anything else about either of these topics, this self-published paper is not reliable by Wikipedia’s standards, and probably shouldn’t be cited there at all. Even if by some stretch of policy it can be considered reliable, self-published sources by definition can’t be very prominent, and it’s giving far too much weight to this source for it to be one of the Wikipedia article’s two most-cited sources.
A Troublesome Inheritance – This book by Nicholas Wade has received a lot of attention since its publication two months ago, but Wikipedia’s summary of the book and its reception are sorely lacking. One omission stands out: from how its reception is summarized in the Wikipedia article, one would assume the book’s reception was entirely negative. In fact, while reviews of the book have generally criticized Wade for not adequately supporting his narrative of how race differences affect society, several of its major reviews have been supportive of Wade’s explanation about the biological meaning of race. This omission is especially noticeable if one looks at the quotes in the article’s “references” section, which selectively quote some of the reviews in order to only include the negative parts. For example, based on this quote from the New York Review of Books, you might assume the review being quoted did nothing but criticize the book. But here’s an example what the same review also said, and that that the Wikipedia article leaves out:
Only rarely do all Europeans carry a genetic variant that does not appear in all East Asians. But across our vast genomes, these statistical differences add up, and geneticists have little difficulty concluding that one person’s genome looks European and another person’s looks East Asian. To put the conclusion more technically, the genomes of various human beings fall into several reasonably well-defined clusters when analyzed statistically, and these clusters generally correspond to continent of origin. In this statistical sense, races are real.
A similar quote can be found in the in The Washington Post review (which isn’t cited in the Wikipedia article, but probably should be).
He musters a good deal of persuasive evidence that, as he puts it, “human evolution has been recent, copious, and regional.” It stands to reason that this should be so. After all, as evolutionary biologists have shown for countless species, the geographic separation of a population over time results in adaptations to specific environments. And, for many thousands of years, most human populations were separated into diverse geographic and social settings.
This is another article where you might encounter some opposition to fixing the problems with it, because the editor who’s responsible for these omissions is still active. However, selectively quoting book reviews in order to only include the negative parts is an obvious enough violation of the “neutral point of view” policy that it really shouldn’t be ignored, especially not in an article about a book that’s both popular and recently-published. A not-quite-exhaustive list of the book’s reviews can be found in the post here, although remember that self-published reviews can only be used as sources if they’re written by established experts.
Don’t engage in meta-discussions
The Wikipedia editor Maunus asked me to include this point, because he’s encountered this mistake from newbies especially often. Inexperienced editors often bring up questions or arguments that aren’t directly related to any change or proposed change to the articles, such as arguments about what ideas are mainstream, or questions about what someone thinks of a particular blog. An example of that is this comment by Wajajad. The reason this is best to avoid is because it doesn’t contribute directly to improving the quality of articles, and most members of Wikipedia would rather not have their time taken up with something that isn’t productive. A related policy is Wikipedia is not a forum.
Do focus on identifying specific problems, and suggesting specific improvements
Instead of bringing up these broad-level topics, it’s much more useful to focus on identifying specific ways the articles can be improved. An hypothetical example of an unhelpful comment would be, “I think hereditarian perspective is under-represented at Wikipedia” (followed by the reasons why you think that). A hypothetical example of a helpful comment (on the talk page of the Ashkenazi Jewish intelligence article) would be, “I think some of the citations to Brian Ferguson’s paper should be removed, because it’s undue weight for a self-published paper to be one of the Wikipedia article’s two most-cited sources.” The first post is likely to result in an argument that won’t actually resolve anything, but the second post could potentially result in a problem with one of the articles being fixed.
It’s important for any objections you raise to be firmly grounded in Wikipedia policy. In the “helpful” example given above, the policies being referenced are the verifiability requirement for sources, and the requirement to avoid giving undue weight to low-prominence sources. As another example, if someone were to argue that the biological theories of causation should be added back to the “Race and crime in the United States” article, that should not be based on a scientific argument about these theories. Instead, it should be based on the fact that widely-used textbooks such as Gabbidon and Greene’s book include discussion of these theories, and that “neutral point of view” policy requires all ideas to be included in proportion to their prominence in the source material. Wikipedia policies such as NPOV apply to all articles, regardless of the personal opinions of editors involved in them. If other people appear to be misunderstanding or misrepresenting policy in order to make an article conform to their viewpoint, that’s an example of a situation where it might be a good idea to start a request for comment.
Editors of this topic at Wikipedia have a wide range of viewpoints on the cause of racial IQ gaps, and trying to argue with them about what ideas are mainstream or supported by evidence is a recipe for conflict. On the other hand, it’s a much more straightforward process to determine whether a certain part of an article violates one of Wikipedia’s policies, and this is an area where editors with different viewpoints can sometimes find common ground with each other. If you and another editor agree on what sort of changes are necessary to make an article comply with policy, it should be possible to work together with them even if their personal opinion about race and intelligence differs from yours.
Don’t try to take on the world
Even in cases where you’re unequivocally right, some of the time you aren’t going to get your way. When there’s a large group of editors who disagree with you, and Wikipedia’s dispute resolution processes haven’t found anyone who supports your perspective, it’s important to recognize when to drop an issue. This is especially the case for arguing over small changes that are relatively unimportant. One of the worst mistakes anyone can make in this situation is trying to argue with every single person who disagrees with you, and not only because of how much attention this can attract.
As an example of how costly this mistake can be, consider the editor BeauPhenomene. BeauPhenomene registered at Wikipedia on April 24th, and began trying to make changes to the main Race and intelligence article on May 19th. Trying to change a highly-visible article without much experience was one mistake, and a second mistake was edit warring. The change he was trying to make was only a difference of a single word, yet he also argued aggressively with everyone who disagreed with him about it. As a result of all these things, he was blocked indefinitely after having been involved in the article for only a few hours. (“Indefinite” means that the block has no expiration date.) BeauPhenomene might have been able to contribute something valuable to Wikipedia if he’d known how to conduct himself, but as a result of this mistake he’s probably lost that opportunity for good.
Do build a relationship with other editors, and with each other
The more friends a person makes at Wikipedia, and the more influence they acquire, the more they’ll be able to accomplish there. This principle applies to other editors’ willingness to listen to you, and it also applies to how severely or leniently you’ll be punished if you make a mistake.
For a striking demonstration of the difference this makes, compare BeauPhenomene’s meager editing history to the summary of Mathsci’s editing linked to earlier in this post. Both editors eventually were indefinitely blocked, but Mathsci received that block after doggedly pursuing a dispute for almost four years, while BeauPhenomene was blocked after doing the same thing for only about 12 hours. The period of October through December 2012, during which race and intelligence related battles made up around two-thirds of Mathsci’s participation in Wikipedia for three months, alone accounts for at least a hundred times the amount of battleground behavior that BeauPhenomene engaged in. The reason this difference exists is because new editors have no social status, whereas Mathsci had a very high status in the community, and had earned the respect of many other editors, including several admins.
One of the most important ways to earn respect is by contributing to articles on a wide variety of topics, including those that have nothing to do with race or intelligence. Editors always are taken more seriously at Wikipedia if they appear devoted to improving a diverse range of topics, and the more friends you make on those articles, the more people you’ll have to potentially defend you if that becomes necessary. It’s also important to form a relationship with editors who care about improving race and intelligence related articles, since these are editors you might be interacting with periodically. The following are five editors who might be able to help you improve these articles, if you request their help in their user talk.
Xxanthippe – Xxanthippe was involved in race and intelligence articles several years ago, with a focus on trying to make them as balanced as possible. Although she’s still quite active in general, it’s been a long time since she last edited that topic. I don’t know whether she would be interested in participating there again, but it might be worth asking.
Silver seren – Like Xxanthippe, Silver seren is quite active, but only occasionally chooses to get involved in articles related to race or intelligence. However, when he does choose to get involved, his involvement usually is helpful. Silver seren has recently told me that he seldom participates in these articles because he finds the atmosphere there unpleasant, but that he’s also willing to provide assistance to newbies if they need it.
Dbachmann – Dbachmann is another fairly active user, who seldom participates in articles related to race and intelligence, but whose occasional involvement there is usually helpful. He’s told me he would be happy to assist newbies who are trying to improve articles on this topic. An additional benefit of earning Dbachmann’s respect is that unlike the other users I’ve listed here, Dbachmann is an admin, so he has a little more influence at Wikipedia than most ordinary users do. However, new editors should not expect Dbachmann to directly defend them using his admin powers, because the conduct rules for administrators don’t allow any admin to use their powers in a dispute where they’ve been involved.
Maunus – This is going to be a controversial choice for some readers, because I know that several of you have come into conflict with Maunus when he opposed you about material that he felt was over-representing the hereditarian hypothesis, or that wasn’t reliably sourced. However, what many people don’t appreciate about Maunus is that he generally cares more that Wikipedia articles comply with policy than that they conform to his personal opinions, so he’s been willing to fix problems with articles even if the problems are favorable to his own viewpoint. Editors willing to do this are extremely rare–aside from Maunus, there have not been any such editors active on these articles for the past six months. One recent example of Maunus’s willingness to do this is what while rewriting most of the main race and intelligence article over the past month, he extensively consulted with me about the article’s sourcing and neutrality, and several of the sources that he used for these edits were book chapters I had photographed for him. If you can clearly explain how a problem with an article is violating a Wikipedia policy, Maunus probably will be willing to help you fix it.
Varoon Arya – If you’ve read some of the external material linked earlier in this post, you’ll already know who this editor is. He retired his account in 2010, giving Mathsci’s behavior towards him as one of the reasons, and the main reason the article “Race and Crime in the United States” has fallen apart since then is because Varoon Arya had been the only editor maintaining it. He indicated in 2012 that he was thinking of reactivating his account, but at this point it’s been almost two years since he was last active. I’ve included him in this list because if someone were to inform him in his user talk that Mathsci is now banned, and that the editing environment on these articles has improved from what it was like in 2010, I think it’s possible he would be interested participating there again. This would be a significant benefit for all of Wikipedia’s articles related to race–during the year or so that Varoon Arya was editing them, he was more knowledgeable about the topic than almost anyone else.
It’s also important that you learn to cooperate with each other. There actually are only three or four editors responsible for most of the persistent problems with these articles, and they’re almost certainly outnumbered by the editors who care about the articles being improved. The reason some of these problems are so difficult to fix is because the editors responsible for them know how to cooperate with one another, by monitoring the same articles and supporting one another in disputes. On the other hand, most of the editors trying to improve the articles haven’t generally had this sense of cooperation.
One example of this difference in organization is the attempts various editors have made at removing the violations of BLP policy in the Richard Lynn article. Five different editors have objected to these violations, and only three editors have seemed determined to keep them in the article. However, the five editors who raised these objections were mostly people who had no long-term involvement in the article, and who brought up their objections at different times from one another, whereas the three who stopped the violations from being removed were all involved at the same time and worked together as a group. In situations like this, organization and cooperation make far more of a difference than majority opinion.
Charles Ainsworth (a.k.a. Cla68) has written a blog post at Wikipediocracy explaining how a small group of editors typically goes about controlling a topic at Wikipedia, titled How to control a topic. I suggest that everyone who’s intending to participate in these articles at Wikipedia read this post. I don’t endorse all of the advice given there (some of these are pretty advanced tactics), but everyone who gets involved in a controversial topic at Wikipedia should understand that these are the tactics other people will likely be using.
When editors who oppose you are using these tactics, the only way it’s possible to counteract them is by working together with one another. You should be careful not to abuse this ability. When an article is dominated by a group of editors with the same viewpoint, it’s possible to make the article conform entirely to that viewpoint–I know this because I’ve seen it done. But doing this is a bad idea, and not only because it runs contrary to the “neutral point of view” policy. Just as the articles’ current situation has attracted the attention of people who read this blog, it will attract people with the opposite viewpoint if the hereditarian perspective becomes excessively dominant. If that happens, you will very quickly find yourselves outnumbered.
My hope is that the readers of this blog who are participating at Wikipedia will be able to use this information to become more productive there, while also avoiding the behaviors that are regarded as problematic. Those of you who care about improving Wikipedia’s coverage of these topics have more power to do that than you realize, but you have to learn to play by the rules.