Requests for comment/Authoritarian vibe on the Romanian Wikipedia

From Meta, a Wikimedia project coordination wiki

This is a subpage; for more information, see the Requests for comments page.

Hi there,

This is a help-my-wp-has-been-hijacked-by-tyrants type of RfC, and I am fully aware that this does not bode well.

What might make it different from some other such RfCs is that I haven't been blocked anywhere, and I'm not here to complain that my freedom is being restricted or that some people promote a particular point of view. The purpose of this RfC is to get external opinions on the way some specific incidents have been handled by the admins, in the hope of getting some of them to realize their project may be on a slippery authoritarian slope.

The Wikimedia project in question is the Romanian Wikipedia. I'll start by giving a bit of background on my involvement in Wikipedia, because I think some people might find that relevant. Then I'll try to give my impression of the way the project is being led, with some concrete examples. Finally I will request comments on some specific points.

Although I have floated the idea, I have not yet informed the ro.wp community about this RfC — but I will do so right after I finish typing what follows.

About my contributions to Wikipedia[edit]

I've been a contributor to the English and French Wikipedia for a little over 10 years. I'm a pretty unremarkable editor, by all metrics (en.wp; fr.wp). Most of my edits are to math articles, especially on en.wp, and are are either corrections of errors/typos, or small clarifications.

I do not interact much with other users (18% of my edits are to talk pages on en.wp; 10% on fr.wp), and I could hardly be accused of looking for conflict or of making controversial edits on these projects (0 reverted edit on en.wp; 3 on fr.wp, and I made some [if not all?] of those myself).

For a full disclaimer: I've been involved in two minor conflicts in the past (here, continued there; and here — my interlocutor decided to delete the rest of the conversation). None of them led to edit-warring, and both resolved fairly quickly on their own.

In August 2019, I created an account on the Romanian Wikipedia, but became active there only in June 2022. My intent was to do the same kind of edits as on en.wp and fr.wp, namely to correct errors in math articles. That did not prove possible, for reasons that are beyond the scope of this RfC, and as a result I came to interact more with other users (34% of edits on talk pages and 16% on WP) and to familiarize myself with the administration of ro.wp.

Personal impression[edit]

(this section does not contain many concrete elements; the goal is mostly to lay out my concerns, and what I expect from this RfC — some might consider it a rant)

The first thing that shocked me is the extremely small number of editors, even when considering the number of Romanian speakers. Of course, different languages have different cultural influences and as a result it is hard quantify this / know what to make of it. However, I think the order of magnitude of the figures is worth having in mind.

There are about 275 monthly active editors on ro.wp vs ~5.6K on fr.wp. If we look at this from a per-L1-speaker basis, that's 12ppm vs 70ppm. And if we look across all speakers, then even taking the most generous estimates for the number of French speakers (450 million), we get 10ppm vs 12.5ppm. Bearing in mind that more than 70% of French speakers are in Africa, but that about 10% of the active editors are there, this is hardly a fair comparison; and yet it still suggests a lower involvement of Romanian speakers compared to French speakers.

My point here is not to do original research or to draw hard conclusions, but simply to say that the feeling I got that there is an unusually low engagement of Romanian speakers with the Romanian Wikipedia seems credible. At first, I did not think much of it. However, after more than one year contributing to the project, I came to the conclusion that the authoritarian way that the project is administrated most likely plays a part in this. As a concrete example of this, the specific examples of abuse described below led two math editors to leave the project (one of them being me).

I have tried to voice this opinion — at first, in a toned-down and conciliatory way; then more explicitly and vocally. That was spitting in the wind: all I got was what I interpret as attempts at intimidation, in particular by one administrator whose abusive behavior I had pointed out (e.g, formal complaints that led nowhere or absurd and disruptive edit-warring) and disparaging comments from some editors (e.g, "Batman... Batman...").

Recognizing a brick wall — and not intending to keep yelling at it — I made the decision to leave ro.wp. I know some will interpret this RfC as me slamming the door on my way out; the reality is that I would genuinely want ro.wp to thrive (my love of the Romanian language being one of the main reasons), and that in my opinion there are many good editors (and even administrators) there. This RfC is aimed at them, in the hope that external input might force them to reconsider some of the behavior that has become normalized — and hopefully fix a few things.

Concrete examples of abusive behaviors[edit]

(note: in what follows I will use "F" to refer to this user; this is because they claimed to be young and to be using their username on other parts of the Internet — so I want to limit the amount of litigious content that pops up when someone Googles their username. Other users I will mention are longtime Wikipedia admins, so I think that makes their online personas public figures)

Abusive behavior takes many forms on ro.wp, ranging from (relatively harmless, but very common) questionable reverts by admins to textbook examples of abuse — such as admins that take care of formal complaints filed against them before other admins have had a chance to look at them... and other admins refusing to condemn it when faced with the facts. I will not try to make an extensive list, but I will simply try to give one example of each type of abuse I witnessed from admins (or former admins). In rough order of severity (and thus starting with things that are not abuse):

  • Questionable reverts: not an abuse, not even a big problem; but something relatively common which I think plays a role in filtering out users.
It's hard to pick a specific example — because I obviously didn't keep track of such minor things — but I happen to remember this revert. Here, the edit was made by a valuable editor. For the most part it was an improvement (fixed 4 typos, improved a formulation, added 3 legitimate wiki-links); however it contained inter-wiki links, which are discouraged (and one of them was, to tell the truth, was not necessary). Instead of fixing things by simply removing the inter-wiki links — we're talking seconds of work here — an admin just decided to revert the whole edit, thereby reintroducing the typos, etc.
In many cases, these questionable reverts can be interpreted false positives for vandalism. But when the author of the reverted edit is a long-term valuable contributor and when the revert reintroduces typos/errors, it can feel like the person who did the revert has little appreciation for the work of others.
  • Strange peculiarities of the official guidelines: again, not an abuse; just one of the quirks of ro.wp that sets the tone.
Most of the official policies are translated from en.wp, with some interesting differences. My favorite one — which in my opinion has a totalitarian vibe to it — is Epuizarea răbdării comunității ("Exhausting the patience of the commmunity"), an official rationale for blocking users. It translates to "When, through their behavior, a user exasperates the community (and that their actions are not covered by another provision of this set of rules), it is allowed to ban them.". But not to worry, there is not risk of abuse: "The administrator who bans the user has to make sure the community agrees"!
In conjunction with what I was saying about the community being extremely small — and considering that most of the admins know each other off-wiki — it can clearly serve as a carte blanche to ban people without valid motive.
Interestingly, that rationale is sometimes invoked by IP users who seem to know the official guidelines pretty well and to be constantly lurking on the talk pages — but none of the admins seem to think there might be something a bit fishy going on here.
  • Admins refusing to follow official guidelines: the Romanian WP:BC is more or less a translation of the English WP:GF. As in the English version, it has the following provision: "Totuși această politică nu cere utilizatorilor să continue să presupună buna credință odată ce apar dovezi extrem de solide în sens contrar." ("Nevertheless, this guideline does not require the users to assume good faith when there is extremely solid evidence to the contrary" — the emphasis is not mine).
For reasons that kept changing as they were proven wrong, the newly registered math editor F was accused by the administrator Strainu of being the disruptive IP user 178.138.* (banned on en.wp and ro.wp). The timeline of the accusations would be comical, if not sad: first the user was accused of inventing sources (when a simple query on Google would have shown this is not the case); then they were accused of having created their account right after the ban of 178.138 (when checking the date of creation of the account would have showed it was created one month prior to the ban); then they were accused of having started to edit articles shortly before the ban (which... OK???); then they were accused of editing the same articles as 178.138, (which is incorrect: 178.138.* edited articles on many topics, with a focus on high-school level mathematics, whereas F only edited university-level math articles)... In the end, the definitive piece of [sic] extremely solid evidence was universally agreed by the admins to be the "unmatched experience and patter-recognition abilities" of the admin.
My objection that this could hardly be considered evidence, let alone "extremely solid" one did not matter. Neither did the many concrete pieces of evidence that, in fact, suggested that F was not 178.138.* (quality vs disruptive edits; completely different level of understanding of mathematics; different use of some special characters such as the Romanian „ ”; etc — I can detail and provide concrete evidence if needed, but I think that is not the point of this RfC).
Despite being repeatedly pressed to provide concrete evidence to justify their actions, Strainu explicitely refused to do so and kept harrassing F — repeatedly calling them a "LTA" (long term abuser) and adopting various abusive behaviors are detailed in separate points below.
  • Admins using disruptive edit-warring to intimidate: new we enter "inappropriate behavior" territory. An example can be seen here: after deciding that F was the IP user 178.138.*, Strainu actively "hunted" them, making gratuitous disparaging comments and reverting/reworking several edits, including on advanced topics (e.g, commutative algebra) that this admin does not appear to be familiar with. All of this user's edits were legitimate, and all were clear improvements (except for one, which contained a typo in a formula). Thus, unsurprisingly, by doing so the admin introduced some problems an article: in the article on primary ideals, they introduced some weird redundancy in the definition — the equivalent of saying "let x be a positive number such that x > 0", for people familiar with the subject. I reverted this, explaining the problem. The admin reverted without explanation. I reverted, with more detailed explanations in the edit summary and on the talk page. The admin then invented a "conflict of edition" between F and myself — something that both of us have refuted — and warned me I could get banned. Result: the article is locked in a suboptimal state, and the incident contributed to two math editors leaving the project.
  • Admins issuing abusive warnings that can then be used to ban users: on ro.wp, news users are supposed to get a warning for each disruptive edit, and get banned if they another disruptive edit after reaching the warning level called "singurul avertisment". In theory, warnings should be tied to specific edits, and should remain polite; the first few ones should also assume good faith.
Shortly after his first contributions — which clearly showed their good intentions and a university-level understanding of mathematics (I can detail if needed), the user F received, without any prior warnings the "singurul avertisment", with the motive "NU INTRODUCEȚI SURSE IMAGINARE!" ("DO NOT ADD IMAGINARY SOURCES!"). The edit that triggered the warning was not linked, but it was this one. Simply typing the titles of the sources on Google showed they existed, and were relevant.
F politely objected, providing the sources. The warning was not removed. F then filed a polite, concise and well-argued formal complaint. Nothing changed. I politely contacted the admin on their talk page, to explain I thought they were mistaken. Nothing. I tried to get the attention of the community. Nothing. I filed a formal complaint: at last, one of the admins ruled that the warning should be removed. The warning was not removed — instead, its justification was changed to "RESPECTAȚI ÎNTOCMAI CELE AFIRMATE ÎN SURSE!" ("RESPECT IN ALL POINTS WHAT IS AFFIRMED IN THE SOURCES!"), with no link to a specific edit. Since there was a typo in one of the edits made by F (and one only — at this stage, F did not have many contributions and they have been scrutinized by Strainu [and to some extent by myself]), one can speculate that this is what is referred to. Under the strictest interpretation of the guidelines, this could justify a level-1 warning, with a friendly message. So I asked for this to be done — which brings us to the next point.
  • Admins take care of complaints filed against them: one example here. I do not think there is much to be added here — except that I filed a second complaint to say I thought there was something wrong when the judge was the defendant. No one admitted this — the best I could get was one person to say they thought it was "o greșeală" ("a mistake"). Most admins simply did not comment; and the people who commented mostly praised the admin and made fun of me ("Batman... Batman..." or "purificatorul proiectului chemat să-i corecteze disfuncționalitățile cu Gladius Dei în mâini" ["the purifier of the project called to correct its dysfunctions with the Gladius Dei in hand"]).
  • Threats to the careers of editors: in my opinion, this is more serious than the previous point, which is why I put it last. However, it's also a less clear case of abuse, because the threats are not explicit.
An example can be found here: "Mă mir de tonul dvs. și al lui F[...]. Aceste discuții sunt publice, iar angajatorii pot înțelege că cei care discută sunt inflexibili. Poate pe dvs. asta nu vă mai afectează, dar pe F[...] da." ("I am surprized by your tone and that of F[...]. These discussions are public, and the employers can understand that the people discussing are rigid. This may not affect you anymore, but F[...] yes").
This requires some subtext. First, the editor who wrote this claims to be a retired academic in Romania, and they who know I'm an academic outside of Romania — so they have no leverage over me; but F has claimed to be a master's student, and their contributions paint the picture of someone who might well choose to pursue a career in academia. Of course, the probability that someone in charge of hiring people might, on their own, find the Wikipedia username of the person they are interviewing, and then start reading technical talk pages, is close to zero — so what are we supposed to understand?
Moreover, to better understand the above threat, it is also useful to know that the person making it has a history of making such threats on the profesionnal life of editors, e.g, threatening to call employers. For instance, about 10 years ago, this editor — who was an admin back then, and happened to know the identity of one of their "wikinemies" — made the explicit threat to call at their workplace to say that this person had been editing Wikipedia during working hours.
I was not on ro.wp back then, so I do not know exactly how the issue was handled; but it is my understanding that some users feel it was not handled properly.

This sums up the kind of behavior I witnessed during my short involvement on ro.wp. It has led me to conclude that there seems to be a disturbing authoritarian component to the project — and that the best thing for me to do is to leave.


Here is a list of the points on which I would appreciate some comments — but of course comments are welcome about any of the points above. It is normal that:

  • non-admins can close discussions between other editors without justification or approval of admins?
  • admins can choose not to assume good-faith based on their "personal experience" alone, without giving specifics?
  • admins can issue "ejection seat" warnings without previous warnings nor precise motives?
  • admins can take care of the formal complaints filed against them?
  • the admin board simply stays silent when pressed to answer the questions above?

As I said, I am leaving ro.wp so I am not interested in the admins justifying each of the specitic the actions listed above / their inaction upon seeing them. What I am interested in is:

  • [for personal development] knowing if I'm over-reacting or if there is indeed something not right;
  • [for the community] determining how these abuses should have been handled, and what can be done to ensure similar situations would be handled appropriately in the future.

Thanks to the people who have taken the time to read all of this, and thanks in advance to those who will choose to get involved in the RfC. I will have limited access to the internet over the next few days and may not be able to respond until Sunday evening, but I am of course willing to clarify anything or provide specifics if the factuality of my account above is questioned.

Best, Malparti (talk) 15:01, 26 January 2024 (UTC)[reply]


Yes, there are some problems, and admins can get sometimes irritated. But by and large left behind its most serious faults. And of course, leaving such faults behind, meant that some prolific editors were no longer welcome, either they left willingly, or they were banned. So, that explains the small number of active editors. Applying a saying about Mensa members, leading Wikipedians is like herding cats. Wikipedians are difficult to get to respect the rules, and sometimes admins make mistakes, like anyone else.

E.g. while admins do not necessarily endorse such positions, they tend not to offend editors who make propaganda for the nation or for the church, while on that is speedily flagged as a bad attitude. Again: in the beginnings of this was much worse, so has greatly improved. Tgeorgescu (talk) 03:39, 27 January 2024 (UTC)[reply]