User:Imetsia/The new templates and User:Victar's block

On 9 April 2021, Inqilābī started a discussion to propose consistent wording for etymology templates. Eventually, that effort was modified, after consultation with the community, into the push for the {{bor+}} and {{inh+}} templates. The two templates would automatically produce the language “Borrowed from” and “Inherited from” respectively, thus saving users some manual labor. A vote was made to create the templates, its result was disputed, and one user, Victar, decided to revert every instance of the templates’ usage. I blocked that user on two separate occasions for that behavior. This essay comes after Mahāgaja reverted the second block I issued. It intends to argue in favor of the {{bor+}} and {{inh+}} templates, and both of Victar's blocks.

A vote was unnecessary in the first place

The early beginnings of this controversy lie in Victar's manipulation of another user into creating the vote. Victar himself has long argued that this vote is “essentially a rollback of the vote that removed the lead text from the template Wiktionary:Votes/2017-06/borrowing, borrowed, which is why a vote [is] required.” Or, using a slightly different version of the argument, "Having lead text in derivational templates has been a long contended issue, which is why a vote was especially needed in this case." Make no mistake: Victar intentionally advanced this argument to push forward a useless vote and reduce the chances that the bor+ and inh+ templates would come into operation. Unfortunately, Inqilābī was successfully "misled by Victar & Co.’s contention" into running the vote.

Victar's reasoning is plausible at first glance, but look an inch further and problems emerge. The 2017 vote rested on the principle that users should have the flexibility to format the text preceding the template as they wished. PUC notes that “[t]here's a reason why I pushed for removing the text from {{bor}}: it's more flexible that way….” On the original vote, WikiTiki89 again stressed that “[i]t's more work to have to override it than it would be to just type the words ‘borrowed from.’” Yet when it comes to the most recent vote, opponents seem to change their tune. Rather than affording users the choice between manually adding “Borrowed from” or having the template do it automatically; opponents would rather impose their own editing preferences on other users. No one is obliged to use the bor+ template instead of the traditional bor template: It is their choice. If a principal argument is “flexibility,” the obvious choice is to support the creation of the new templates.

Thus, the 2021 vote in no way overturns the 2017 vote. It merely adds another option, while not negating the possibility that users can still use the original bor template if they prefer. This sort of action “add[s] additional functionality without changing existing functionality,” and thus requires no consensus-building.

The infirmities with Victar’s actions do not end there. Just one month before the templates vote, Victar was caught again misleading a user into creating a pointless vote. In the “Merging Prakrit lects into one” vote, Metaknowledge called it a “needless bit of bureaucracy.” They continued to urge Victar not to “impos[e] votes” on others “[i]n the future.” Victar’s strategy here is clear. Victar wishes to force a vote on an issue that does not merit it, and then claim that it “failed” when it does not reach a 2/3 supermajority. They thus impede the passage of a proposal by needlessly subjecting it to plebiscite. Note that even if an overwhelming majority supports a proposal, Victar still wants to stand in its way and offer the veneer of following process and sticking to technicality. That even extends to simple RFD votes that do not require the rigid 2/3 standard.

To succeed on the argument that the templates needed to be voted on, Victar must necessarily swim upstream. It is now longstanding precedent that we do not need a supermajority to back the creation of a page. That standard has been applied to proper entries since the very beginning of our Wiktionary project. Indeed, we generally require a large majority to delete existing pages through the RFD process. That is a presumption in favor of existing entries. A large majority is needed just to overcome that presumption and delete an entry. In general, entries can sustain their claim to legitimacy unless a consensus is reached that they should be deleted. That same rule has always applied to templates and other non-entry pages. Not once in the history of Wiktionary have we required consensus to create a page, and there’s no reason to grant a departure from that precedent for the bor+ and inh+ templates.

After explaining that template creation and deployment does not require consensus, what’s left of opponents’ arguments in favor of holding the vote? Opponents have conceded that templates do not ordinarily call for ballot-casting. But, needing to justify the creation and enforcement of the vote, the last try is to say that “since this template is one in a series of arguably most used templates, any creation of a template that takes over a part of or even the whole function of {{bor}} or {{inh}} should get a vote.” Merely noting the widespread usage of the templates, however, does little. It is a distinction without a difference, and opponents have failed to explain why that matters. And, more to the point, bor+ and inh+ do not in any way overtake the function of the original templates. Indeed, users can choose the old ones or the new ones. No current template replaces any traditional one.

All of that brings to the fact that “there was no point of having the original vote in the first place.” Templates can be created freely and are subject to the same RFD possibilities as any other regular entry.

The vote clearly wins on its merits

On its merits, creating bor+ and inh+ makes sense. It is simply natural that we include the “Borrowed from” language into the template itself rather than ask users to manually add it on their own. It is a “real convenience” to have the text automatically generated by the templates, especially for those who edit etymology sections frequently. The argument in favor of the new templates makes itself, and it convinced 20 users to support their creation, against the comparatively small group of 10 users who opposed it.

Bear in mind again that users who prefer the old way of signifying inheritance and borrowings can still use the traditional templates. Nothing stops them from not using bor+ and inh+. So their only argument can be that bor+ and inh+ are so fundamentally “bad” that no other user should be able to use them either. No such existing argument is persuasive.

One theory to stop the creation of the templates has been that “This seems like templatisation creep; we can and do write it in full when it might be ambiguous, and in the vast majority of cases it isn't.” The argument fails on two grounds. First, even if this is a personal editing preference, there is no reason why other users cannot “write it in full” if they want to and utilize the newer templates. Second, we typically include the “Borrowed from” language in etymology sections even when the borrowing should seem obvious. Even Victar has said that “I always write out borrowed when applicable, even in the middle of a chain.” A great many users follow this practice, and being able to make use of the then-proposed templates would save some manual labor.

“Some have argued that the template is useless because its whole purpose ‘is simply to save keystrokes....’ But so many of our templates do just that. Consider any number of the headword templates by which a user can type, e.g., {{en-noun}} instead of {{head|en|noun}}. As Lambiam pointed out, "Many of the most-used templates are 'unnecessary'" if we just go by this criterion. I find bor+ very useful, and so do a number of other users. It incorporates the "Borrowed from" language in a way that is natural and helpful.” Indeed, I support the template in large part because of how “natural” it is to include the phrase “Borrowed from.” Naturality is a virtue we can pursue for its own sake, and one which clearly hands the victory to the new templates.

The vote gained consensus support

By any measure, the vote gained consensus support. 20 users voted in its favor, while only 10 voted against it. That is a 2/3 supermajority that satisfies even our most stringent standards. The only way to read the vote as a “failure” is by relying on a technicality. One supporter cast his vote just hours after the deadline.

First, a variance should have been granted to allow including the late vote. Not doing so would deny 2/3 of our users the implementation of a very popular proposal. In response, some have argued that a variance would do little because “even though Aryaman was too late with a support vote which would've won over the vote, so was PUC with an oppose vote.” It is a frail response, however, because PUC was editing anonymously at the time and “an IP cannot vote.” In addition, speculation about these alternative universes is improper and unnecessary. We could instead imagine, for example, that Aryaman had cast his ballot just seconds before the vote was closed, precluding PUC’s ability to upset the balance.

Second, even if the technical threshold was not fully met, it is clear that a vast majority of our users support the creation of the templates. They may not have reached capital-c “Consensus” by our strict definition, but they did certainly reach consensus by looser definitions. Those who revert or obstruct their usage are thus running counter to the will of our users.

Third, our rules should be adjustable and understood as general “principles, not laws…. Policies and guidelines exist only as rough approximations of their underlying principles. They are not intended to provide an exact or complete definition of the principles in all circumstances. They must be understood in context, using some common sense and discretion.” Wiktionary operates on the idea that we follow consensus. A vote that has gained overwhelming support should be embraced, not derided on a narrow objection.

Even interpreting the vote as a failure, the templates were still created legitimately

Even if we accept arguendo that the vote did not pass, that still does not foreclose the creation of the templates. Immediately after the vote’s conclusion, a bureaucrat noted that “we don't generally require consensus to create a template; anyone can do so. So the fact that this vote failed with ‘no consensus’ has very little effect; all it means is that no one is required to create and/or use these templates.” Stated another way, the “vote does not carry force to disallow the creation or existence of these templates anyway. The bloody sodding thing almost passed with 2/3s anyway, it is utterly ridiculous to read that vote as a consensus for deletion of the templates.”

Once the templates are created, rightly or wrongly, they exist. The status quo then becomes that the templates are in circulation, and any user can implement them at their discretion.

Victar’s edit-warring earned them a three-day block

Over and over again, Victar refused to acknowledge the status quo and the will of our users. For some time, he reverted usage of the bor+ template, and I did nothing about it for the first few revisions. Once it became clear that Victar was bent on reverting every instance of the template, I asked him to stop doing so. On ten separate occasions, I reminded Victar to keep the templates in place, on pain of a block. Without regard to my warnings, Victar undid the bor+ template. Every time I reverted the page back to the original, Victar met that with another reversion. That clearly constitutes edit-warring.

If Victar wanted to avoid the block, they would have engaged other users in a discussion about the standing of the disputed bor+ and inh+ templates. Instead, contrary to their better judgment, they continued to remove the template from every page on which it was used. Because the status quo was that the templates existed, that votes are not necessary for the operation of templates, and that the pages themselves originally contained the bor+ template; it was Victar’s burden to work to resolve this matter. In the meanwhile, he should have suspended his reversions.

By not doing so and instead forging ahead in his pursuits, Victar showed that he was not interested in cooperative editing. He tested boundaries and tried to see how much he could get away with before administrative action would be taken. That is an issue in its own right and runs counter to Wiktionary policy. On that ground alone Victar could have been blocked.

But the greater problem is that Victar failed to disallow the creation of the templates with the vote, then failed again at RFDO, and nevertheless insisted on chance after chance to reverse community consensus. That path turned out to be a long and lonely one. Despite the fact that the templates existed and were legitimate, Victar took it upon himself to be the one person to obstruct the operation of the template. It was a vivid illustration of the dangers of ceding the disposition of template usage to just one individual. Other users were deprived of the opportunity to put the templates to work. A block was necessary to encourage Victar to end his crusade against the templates.

Otherwise, “[t]he person ‘willing to go through the pain of replacing it with’ the old format was systematically ensuring that the new bor+ and inh+ templates were put out of operation. [Despite] the templates [being] legitimate… one user alone had taken it into his own hands to de facto prevent their usage. It takes us to a situation where the templates might just as well not exist if they can never really be put to use, just because one user obstructs their deployment.”

Victar’s persistence in abusive editing earned them a two-week block

When Victar’s block expired after three days, it appeared that they had not learned their lesson. Indeed, they continued to revert the bor+ template, to the point that another administrator wrote a message on my user page to express concern about Victar’s behaviors. They also called for a longer block to be issued.

I was at first hesitant to mete out a lengthier block for fear of “being too heavy-handed.” I am one of the more lenient, block-averse admins on the site. In the two high-profile blocks in which I weighed in, I argued that Dentonius should have been unblocked and, later, that Brutal Russian should have been unblocked. In my view, we should only block users if they blatantly violate policy and (for established users) if they have clearly been told to desist in behavior harmful to our site. Both criteria hold in this case. In addition, it became clearer to me that a longer block was required after consulting with Benwing2. The block, in other words, was not the result of one user’s interpretation and reflection; but only took place after a concurrence from another well-respected admin. Moreover, it respected our own blocking policy and the time guidelines it sets out.

It is then true that admins should be fair and even-handed in their enforcement actions. One expression of this is following the time guidelines for blocks. Another is that we should enforce the rules whether or not we are happy with the result. In this case, I have enforced the usage of the inh+ template despite the fact that I do not personally like it. This shows adherence to principle and fairness.

Reverting the block was improper and an abuse of discretion

Admins generally have broad discretion to decide when to block and for how long. And for good reason. Admins need to act swiftly and resolutely to stop behavior that hinders Wiktionary's progress. Thus, it is generally accepted that one admin should not unilaterally undo another admin’s block except “in extreme cases of abuse.” Thus, for Mahāgaja to succeed in unblocking Victar, he would have to show that the two-week block was not only abusive but an instance of "extreme" abuse. They cannot shoulder that burden.

To begin with, three administrators (including me) are now on the record as supporting the two-week block. Benwing2 suggested the longer block at the start, and Bhagadatta supported it in another discussion. That is in comparison to just two administrators (Metaknowledge and Mahāgaja) vocally wanting to unblock Victar. If the block was indeed "abusive," that seems to be the minority opinion among administrators. At a minimum, further discussion should have taken place before my block on Victar was reverted.

Both Metaknowledge and Mahāgaja believe an unblock was warranted because the templates failed to reach consensus. But, as explained above, templates do not need consensus to be created; and, even if they did, the templates did achieve supermajority support. At the very least, Mahāgaja’s assertion is disputed, so a unilateral unblock was inappropriate and marks an abuse of discretion.

The steps ahead

If there exists a venue for doing so, Mahāgaja’s decision to unblock Victar should be appealed. A selection of admins should vote to support or oppose the unblock. (Admins only because that is the group we have entrusted with the power to block and unblock users). A simple-majority vote should decide the fate of this administrative action. The standard of review should not be de novo but abuse of discretion.

A vote to nullify the previous one about the creation of the templates should take place. It should hold that the templates did not require a vote to begin with, and their operation is legitimate. A more ambitious proposal would be a vote to standardize the template-creation process, reaffirming that they can be created freely and are subject to the standard RFD procedures.

As a last resort, a full-scale revote can take place. It should include only the {{bor+}} template, however, for reasons of both strategy and policy. This measure is the least desirable because, on principle, neither template should require a vote to be created.

Concurrences

• 𝓑𝓱𝓪𝓰𝓪𝓭𝓪𝓽𝓽𝓪^{(𝓽𝓪𝓵𝓴)} 10:44, 18 August 2021 (UTC)
• I agree with all this, and thank Imetsia for creating this. I'm however not so convinced with "vote to nullify the previous one about the creation of the templates should take place" as I'm insecure that's it can also be a no consensus and then these templates will lose. As so many pointed out, vote was unnecessary. And the new vote was created mainly just to stop one user who kept reverting these templates by referencing the old vote. The block on that user should be restored and the templates, kept, as RFD failed to delete it. The old vote's title was also wrong, it should have been "Bot operation to use {{inh+}} & {{bor+}}". This new vote is a risk: we may lose the templates for a long time. But OTOH, the opposition may be crushed. I'm not very sure we should take that risk. I'd instead prefer a peaceful discussion with the opposing people, but it is highly unlikely that they'll listen to us and be convinced to not revert the new templates. They bring the same argument again and again, even if one shows them a thousand times that it's completely invalid. For this reason, I'm okay with the new vote. I am hoping that the 20 who supported the last one support this one also, and this would get a little more support by some other users. Again, thanks a lot Imetsia, for creating such a detailed page with references and all. Svārtava² • 03:17, 23 August 2021 (UTC)
• ·~ dictátor·mundꟾ 11:42, 23 August 2021 (UTC) Any sane being would agree with the points.
• Llittleserie (talk) 07:18, 25 August 2021 (UTC)