User:Imetsia/The state of the plus templates

''This essay is a continuation of User:Imetsia/The new templates and User:Victar's block. Please refer to that essay for background information and previous arguments.''

For something so seemingly simple, few would have thought that the plus templates could have generated so much controversy. By now, the situation has cooled down significantly, but it took a great deal of discussion, followed by two votes, a block, a counter-block, and other measures before this happened. Some of the pivotal moments have already been discussed in a prior essay, but greater emphasis should be placed on Mahagaja’s unblock of Victar. In many ways, it was the defining moment of the templates debacle, when Mahagaja seized the reins of power to exceed the proper limits of his authority. Thereafter, the standing of the plus templates became hopelessly confused.

Mahagaja left it to others to sort out the problem that he had created, putting the onus on template-supporters to decide what strategy could put the situation back in order. A number of possibilities were open: appealing the unblock, revoting on the plus templates, improving the traditional ones, or nullifying the first vote. All of these options bore their benefits, but nullification held the best prospects. The alternatives were too risky. An appeal would have needlessly agitated users and caused them to confuse adjudication with legislation. Revoting failed both on principle and strategy: It would have dignified the original vote and carried virtually no chances of success. Improving the traditional templates could have been a good idea, but the Beer Parlour discussion about that suggested that the proposal was also destined to failure. That left only nullification as a worthwhile attempt, reasserting that templates do not need prior approval for creation and giving us a practical opportunity to keep the templates in place.

Inveterate template-opponents were quick to oppose nullification, manufacturing all sorts of arguments against it; none of them are valid. Some opposers could not be bothered to even defend the validity of the original vote. They only brought themselves to vote against after speculating about the motives for nullification, getting the facts wrong about the first vote, and misattributing blame. Others proposed that the templates’ implementation be cut back to the maximum extent possible so that the plus templates may exist in name only. Many opponents of nullification, including the ones to which I allude above, agreed that the templates were useful, but nevertheless insisted on standing in their way. Ultimately, they were able to cobble together enough of a following to tank the nullification vote with no consensus.

With the failure of the nullification vote, the state of the plus templates once again hung in the balance, leading to other efforts to reaffirm their legitimacy. One approach was continuing to use the templates in disregard of the result of the nullification vote. It was hard to see, however, how this path could lead us to victory. Another idea was a stipulation that could gain us a toehold and achieve some basic demands. I opted for the latter, although disagreement within our camp also made this idea ineffectual. Surprisingly, that did not matter. Victar eventually ended his abuse against the plus templates, a state of affairs which has continued up to the present day.

If there is one lesson to take away from the clamor of the plus-templates controversy, it is that we need a better way to handle dispute-resolution. All signs point to an arbitration committee as our best hope.

Mahagaja’s unblock was egregiously wrong
A single administrator decided, of his own accord, to impose his own personal judgment on our community of editors and, in the process, place the standing of the plus templates into serious doubt.

If Mahagaja had consulted with other administrators and reached a consensus about how to arbitrate in the dispute, his actions might have been tolerable. For some, it might still have been difficult to stomach. Certainly, I and others would have strongly disagreed with it. But everyone is required to follow the democratic will of our community and stand by its final holdings. At that point, users should have treated the matter as settled, set aside their personal feelings about how to best resolve the problem, and move on to the more important task of building a dictionary.

Mahagaja, however, set off on a different course. He decided to bypass the traditional process of notice, discussion, and comment that underpins so much of our governing practice on Wiktionary. There are actions that individual administrators can do on their own, using some discretion and common sense. After all, “[a]dmins need to act swiftly and resolutely to stop behavior that hinders Wiktionary's progress.” Reverting the reasoned decision of another sysop does not fall into this category. Disagreement with another admin’s judgment is acceptable, but actions intended to undermine another admin’s decisions must first be approved by the community. Here, the belief that Victar should have been unblocked was the minority view. Transforming that dissenting opinion into the final, binding resolution between the parties was the wrong course of action. For all the bluster about debate and consensus for the templates, it is telling that Mahagaja et al. have stayed quiet on why they did not seek confirmation before subverting another admin’s decision.

As a substitute to notice and comment, Mahagaja and Metaknowledge tried mightily to defend their position by resting on one sole argument: That the templates “were created in defiance of consensus”. Try as they did to make the argument work, however, the mantle of “consensus” does not fit their claim. The templates’ creation and implementation never required consensus in the first place, and even if they did, the templates did gain consensus by all available definitions. But assume, for the sake of argument, that the vote failed. Even then, the templates’ deployment was legitimate. So, after deploying every rhetorical and political resource at their disposal, Metaknowledge and Mahagaja still come short – no less than three times over – in justifying their behavior against the templates.

Making matters worse, Mahagaja’s unblock is virtually unreviewable. Undoing the unblock would likely be considered combative and an instance of “wheel-warring.” So a unilateral decision based on an unpopular and objectively wrong gloss on our policies must stay in place; we are forced to let it lie based on the arbitrary fact of which administrator was the last to review Victar’s conduct. Not only is this an abuse of discretion; it is an abuse of policy – a means to weaponize existing guidelines in a way that favors one party and disadvantages another.

Stacking the deck in favor of one group of users, however, is not the most fundamental problem with Mahagaja’s behavior. Beneath the surface, the greater issue has to do with humility; no administrator on his own possesses the special wisdom to ordain for everyone else what policies they should consider just or correct. As a result, no user should have the power to annul, all on their own, an antecedent consensus. But in unblocking Victar, this is precisely what Mahagaja did. He short-circuited our bedrock custom of notice and comment on his own theory of why the templates’ creation was illegitimate. Nothing in our policies allows that outcome, and nothing in common sense can sustain it.

Mahagaja’s unblock set the stage for uncertainty about the templates
Through the unblock, Mahagaja had taken it upon himself to pick the winners and losers of the templates situation. According to him, the templates should never have been created, and Victar was right to obstruct their operation. (Never mind that this idea was rejected two-to-one in the previous vote). The unblock effectively held that anti-template partisans were correct, and that no one may use the templates in the future – for if they did, Victar was within his rights to revert it as he had been doing before. Indeed, Victar was emboldened to do so and trumpeted his vindication. Little discussion took place before handing down this verdict. Instead, just one user decided he knew best and could impose it on everyone else.

If there were one upside of unblocking Victar, maybe it could have lain in finally putting an end to the templates situation. In a way, a declarative holding should have extinguished ongoing debates and settled the legitimacy of the plus templates once and for all.

But the opposite happened. The impression was clearly that Mahagaja had acted unfairly and one-sidedly. He acted not according to principles of justice and the greater good, but with prejudice against the templates. Thus, the final decision could not be trusted. After having “con[ned] another user into creating the first vote [and] convince[d] so many to shift the burden of proof for the… templates,” Victar had simply lured an admin to bail him out. The conclusion was unsatisfying because it was illegitimate. It was only natural that users would continue to fight for the templates and implement them in their edits. It could not be that “trickery” and cunning would be enough to frustrate governance on Wiktionary just so that one user could get their way. Rather than putting to rest open questions about the implementation of the plus templates, then, Mahagaja’s decision to take matters into his own hands deepened the uncertainty about the templates’ standing.

The question then became how to challenge Mahagaja’s heavy-handed administration of policy. The unblock had opened the door to greater confusion and new ideas on how to reaffirm the operation of the templates.

Appealing the unblock was not the best option
The most direct approach to undoing Mahagaja’s judicial activism would be to appeal the unblock in the Beer Parlour. As straightforward as it may seem, however, this proposal has a number of problems.

First, it would further stir the passions of users on both sides and channel their energies away from the core issue of the plus templates. To an extent, any other option at the disposal of template-supporters would have kindled infighting among editors. Appealing the block, however, would have amplified tensions even more than the other options would. Contributors would have narrowed their attention to just a handful of personalities on the site (me, Victar, and Mahagaja), discussing the merits and demerits of their conduct. The tenor of these discussions would have been particularly strident, because many would have viewed the move as a vindictive tactic that aimed to bring one user or another into disrepute. Thus, as the discussions went on, users would have moved further and further away from the central issue of the templates and instead attack the errors and follies of other editors.

Second, we lack the mechanism to make any unblock appeal workable. As it is, anyone is allowed to participate in Beer Parlour discussions. If the unblock were a strict affirm-or-reverse vote, all users would be allowed to cast their ballot for or against the unblock. That is problematic because only administrators should be able to adjudicate on such matters. Only administrators are familiar enough with enforcement actions to rule on their appropriateness. Moreover, the argument has been raised that only uninvolved administrators take part in the voting. Putting any of these recommendations into place (or deciding not to) requires another host of separate discussions about whether to limit voting eligibility. To my knowledge, such restrictions are unprecedented for Beer Parlour votes. We simply lack the precedent or a preexisting infrastructure for such a vote to take place successfully.

Lastly, an appeal effort would tempt users to legislate from the bench. On appeal, the only question should be whether blocking Victar comported with our policies. The question is decidedly not whether another admin would have come to a different resolution, or whether a block was objectively the best decision. If we were to overturn every enforcement action because it was not the absolute most correct, our governance would suffer from “the mischievous effects” of mutability. Besides, these are inherently value judgments that legislators can make, but not judges. But it is not a given that other users would adopt the same view. On the contrary, voters would be enticed to view the facts of the case de novo and impart their own personal preferences into their judgment. They would exceed the limits of judicial authority, and we would risk being issued the wrong ruling.

Revoting on the plus templates was not the best option
Another idea might have been to revote on the templates, an option originally advocated for by Inqilābī.

This idea fails as a matter of principle. Templates have never required prior approval for creation, and it should have been no different for the plus templates. Creating a revote both flouts our standard rules for page-creation and dignifies a sham vote, which in turn forms precedent and encourages similar dishonesty in the future.

Furthermore, a hasty revote is unadvisable. Several users have staked out the position that any revote, counter-vote, vote reversal, or nullification is flatly impermissible unless delayed. Such arguments “elide[] a useful distinction between an effort to overturn a decision on its merits and one to reconsider based on a procedural question.” But it does not change the fact that legion users have already convinced themselves that they must presumptively oppose revotes in any form. To be clear, a vote rehashed and reproposed immediately after its conclusion is generally to be resisted. A revote, as distinguished to the vote nullification discussed below, would have been soundly defeated – and for good reason. Even those in favor of the templates should, on principle, vote against a repetitive vote put forward in quick succession.

Improving the original templates might have been a good alternative
While the two prior suggestions can safely be waved off, there was another possibility that had some merit. As PUC and later SGConlaw proposed, template-supporters could have instead started discussions about whether to alter the bor and inh templates themselves. They could include by default the wording “Borrowed from…” and “Inherited from…” respectively, so that we could address the heart of the matter rather than toy with solutions around the margins.

The only real defect in the plan to reword the traditional templates is that it might have been a difficult proposition. Doing so would truly have necessitated a vote, because it would in fact amount to reversing a prior consensus. Given that a previous majority had supported stripping the “Borrowed from” language from the templates, posing a revote on the same question might have been ill-fated. More relevant is the discussion that triggered the entire templates-debate back in April 2021, where a phalanx of users emerged to oppose changing the bor template.

Then again, standards on Wiktionary have changed, and including “Borrowed from” has become mainstream. The list of opposers in the Beer Parlour discussion might have represented a particularly vocal minority. Everything considered, it is not so clear which option would have held greater promise. It was essentially a toss-up between changing the base templates or allowing the newer ones.

Nullifying the plus-templates vote was a necessary and promising measure
The nullification effort was born out of necessity. Mahagaja’s unilateral intervention in Victar’s block was an example of an administrator taking sides rather than interpreting our policies impartially. As wrong as it was, however, it laid down the law that the templates’ deployment was forbidden. Shortly thereafter, those who had engaged in a weeks-long crusade to block the templates crowed about their vindication. They proclaimed that the original vote was now binding law, and its purported failure foreclosed efforts to deploy the plus templates. Absent a clear annulation of the vote, it became clear that template-attackers would continue to repeat the party line against the plus templates. So long as Victar could turn around and resort to the old argument about the vote’s so-called “defeat,” they could keep getting away with obstructing the new templates. Worse, Mahagaja’s preferential use of the unblock power validated Victar’s manipulation of the rules and signaled that others could engage in “all manner of edit-warring” just “to get [their] way after an unfavorable consensus.”

So Mahagaja had thrown the entire templates situation into confusion, picked favorites to advance his own interests, and left it to other users to clean up the mess he had made; now the only question was how to undo the damage. The most natural attempt was to hollow out the plus-templates vote and undermine its central holding. It was time to make “clear that [the] templates vote was a procedural nonstarter,” and that the two-to-one result in its support could not “be interpreted as a decision to strike down the… templates.” Anything short of this resolution would keep the original vote in place and open the door to even more gamesmanship tactics.

Perhaps, some have argued, a simple discussion could have fared better than a new formal vote. Metaknowledge, for example, has made the point that users could have “work[ed] this out like adults (that is, with a discussion), instead of… creating bureaucratic counter-votes.” However, an informal discussion would leave the outcome of the original vote untouched, so that Victar and others could continue to reference it in defense of their actions. They could continue to brand it as black-letter law and claim the supremacy of formal votes over informal discussions. Merely getting into a dialogue with template-attackers was not a solution. Any measure that would allow them to find refuge in a narrow, creative interpretation of the templates-vote would not do.

These concerns are not theoretical. For one thing, everyone should be familiar with Victar’s extraordinary maneuvers to frustrate our governance. He has endeavored time and again to scuttle proposals he personally disliked by inventing novel procedural restrictions. There is no telling what sort of machinations Victar could engage in in the future to try to override community consensus.

Practical considerations also counseled in favor of nullifying the original vote. A number of users such as Gnosandes, Rishabhbhat, and Taimoorahmed11 (who had not previously been active in the discussions) had come out in favor of the new templates. If everyone who supported the templates the last time, plus all the new supporters, would have cast their ballot in the nullification vote, it would have easily passed. The original vote already gained two-thirds support, so any new supporter would simply widen that margin of victory. Alternatively, if some of the prior supporters chose not to participate in the new vote, it should be expected that an equal number of the prior opponents would do the same. There was no reason to think that the opposition was any more or less galvanized than the supporters. On the other hand, there was reason to believe that we could increase turnout even more. If we could direct other casual editors to the vote, they would see for themselves the merits of the plus templates. In addition, they would be swayed by a bandwagon effect after seeing the wave of users voting disproportionately in support of the templates. I tried to do just this by messaging uninvolved users (most of them below the level of autopatroller) to get out the vote. The strategy, however, did not pan out.

One last hope was that previous template opponents could be principled about the nullification effort and vote to support it. Even if they personally disliked the templates, they could still acknowledge that the original vote for their creation was illegitimate. Despite their passions and preconceptions, users could hopefully prioritize the fair and even-handed application of our rules over partisan interests. Administrators, just as much as any other user, should not be searching for expedients to win the day’s case. Instead, they should assess the matter dispassionately and make their best judgment. Sadly, this did not happen. The final vote was decided largely along party lines.

None of the objections to nullification are convincing
Critics of the nullification effort advanced all sorts of arguments to preserve the plus-templates vote. Their justifications lay bare the core reason for nullification: They could not muster a reasonable defense for keeping in place an illegitimate vote that runs counter to policy. Opponents only managed to support their position by attacking the concept of nullification per se, while faulting supporters for failing to meet burdens that do not exist – such as needing to postpone the vote for its own sake.

I will respond to some of the more prominent arguments presented.

"I don't have terribly strong feelings about the templates; at this point, I might actually be slightly in favour of their existence. However, I do have strong feelings about the idea of 'nullifying' a vote to get one's way after an unfavourable consensus. The solution to a poorly conceived vote is not yet another poorly conceived vote. Go work this out like adults (that is, with a discussion), instead of edit-warring, blocking, and creating bureaucratic counter-votes." (Oppose template removed).

Metaknowledge’s response starts off well by announcing that he is now in favor of the templates, but even this fact is not very comforting. Metaknowledge was the very first person to cast an oppose vote for the creation of the templates, even though he now agrees with “their existence.” If only he had voted to support (or even abstain) the last time around, the templates’ legitimacy would be beyond debate and the entire fiasco would have been avoided.

The response only gets worse. Metaknowledge then went on to claim that the nullification vote was just another attempt to forcibly “get one’s way.” To the contrary, annulling a violative vote “offer[ed] the community” a chance to revert its past error – it was “expressly not… an attempt to change the outcome of a vote because you don’t like it.” In other words, there was more to it “than clearly establishing… that the previous vote, by its failure, carrie[d] no weight. It also aim[ed] to establish, as a principle, that creation of a new template does not require prior approval…. So… the [original plus-templates] vote itself was improper – it should not have been held in the first place.” In addition, the new vote addressed the concern that “we lack[ed] a mechanism” to avoid unnecessary votes. Any of these reasons, as well as any listed on the previous essay, are grounds enough to abandon the original plus-templates vote. But Metaknowledge refused to vote to invalidate it, not because he disliked the templates and not because he could really defend it, but because of a dogmatic insistence on opposing “counter-votes.”

If that were not enough to show the absurdity of Metaknowledge’s view, try to take it to its logical conclusion. If nullifying a vote sets a bad example, consider the precedent that would be established if we failed to nullify. It would send a message that any user who “disagrees with… a community decision… [is] free to obstruct its operation,” employing one stratagem after the other “until they are happy with the result. If anything, it would be this… user,” not the nullification-supporters, “who would have their way with our policies ‘just because [they] disagree with the outcome.’” This is supposed to be Metaknowledge’s greatest fear (and the sole motivation for his oppose vote), yet his position leads to the very situation that he professedly wants to avoid.

But even that is not the worst of it. The next part of Metaknowledge’s apologia is that the nullification effort was simply in protest of “an unfavourable consensus.” This is false. “Not even the most dyed-in-the-wool template-opponent can argue” otherwise. The original vote “got 20 support votes against 10 oppose votes, [though Metaknowledge] and others continue to read [it] through a mirror.” So Metaknowledge moved from one dubious claim to another that is even more indefensible. Maybe if read at the right angle, one could fall for his previous arguments, but this one is refuted wholesale by the factual record. At the absolute most, one could claim the original vote had no consensus. But even this extreme interpretation cannot avail Metaknowlege in this case. “That some have taken it upon them to ‘interpret’ a result barely short of a supermajority as a consensus to delete is deplorable and wikilawyering at its worst.”

The myriad problems with Metaknowledge’s opposition do not end there. The next sentence accuses me of “edit-warring, blocking, and creating bureaucratic counter-votes.” To begin with, Metaknowledge’s disapproval of edit-warring is misplaced. Before blaming me for it, perhaps he should have asked who started edit-warring, defied countless notices to stop reverting edits, and acted against the status quo. The answer to all those questions is Victar. For all those reasons and more, “a block was necessary to encourage Victar to end” his spree of harmful edits. Moreover, as argued above, the nullification vote was a necessary measure. And lastly, it is quite bold for someone to take a stand against “bureaucracy” when they actively promoted it in the last vote. If Metaknowledge were serious about cutting red tape, he would have called to halt the plus-templates vote as unnecessary. Not doing so is a testament to just how intermittent his commitment to diminishing bureaucracy has become.

No one would have thought that undoing a violative vote would be so difficult or controversial. Indeed, not even the early-bird opposer to nullification can present a colorable argument to keep the illegitimate vote in place. Instead, Metaknowledge skips from one feeble claim to another, finally landing on a narration that is radically divorced from the actual history of this case. All the while not even pretending to defend the validity of the prior vote. And for what? To prevent us from having two very popular templates and to fuel a peremptory opposition against the concept of a revote.

"As I've stated before, I'm not against the templates per se: if people want to save keystrokes, let them do that. But I think having two sets of competing templates is bad, and here's what I think we should do to avoid that (I'll be quoting myself): '1) the use of bor+ and inh+ would be tolerated in new instances (i.e. when they add a precision that wasn't there before); 2) they could not be used to replace the old templates when these are already preceded by 'Borrowed from'/'Inherited from' (i.e. when it has no impact on the users' end); 3) on the contrary, they could be replaced by the old templates by whoever should wish to do so. They would be in a state of permanent semi-deprecation, if you want.'

Apart from the fact that the new templates create a link to the glossary - something which I'm not convinced is necessary - I'm yet to see a cogent reason to oppose this; all I see is a weird attachment to the templates themselves. As Imetsia said, 'There is no race to see how widespread the templates can become, or if one template can overtake the other'; so why should it matter that the new ones are systematically being replaced by the old ones, as long as one can use them when they want to save some typing? But since the templates defenders seem to be opposed to that idea, I'm forced to oppose. Besides, I agree with Sgconlaw that rather than arguing about this, we should be discussing the underlying issues….'" (Internal links and oppose template removed; arguments to which I have responded elsewhere removed).

PUC’s response is much more measured, but it still has problems of its own.

Like Metaknowledge, PUC begins on a good note. He accepts that one of the templates’ main objectives, automating a task to “save keystrokes,” is laudable. But he does not really rest on that argument. Despite agreeing with the crux of the plus templates, the rest of the response is devoted to impeding the templates’ deployment through a scheme of “semi-deprecation.” To vote as he does, PUC must start from a good premise – that the templates fulfill a legitimate user need – and then muddy the waters. The first step is to say that “having two sets of competing templates is bad.” To begin with, however, the plus templates do not at all compete with the prior ones. “If anything, [they] are complementary,” because anyone can “choose the old ones or the new ones” to their liking. Ideally, the plus templates would be used when it is desirable to produce the “Borrowed/Inherited from…” pre-text in the etymology, while the traditional templates would be used when other pre-text is more appropriate. For those editors who believe it is more natural to have the text produced for them, the plus templates serve a noble goal. For others who want greater leeway in formatting etymologies, they can use the traditional templates. The two sets of templates need not rival for dominance so that one “can overtake the other.”

The next move is to stake everything on “consistency” in application of the templates. Ironically, this claim is itself inconsistent with what PUC had argued in the past. In the Beer Parlour discussion on standardizing borrowings and inheritances, PUC railed against making the templates more uniform. But when given an opportunity to compromise by allowing the plus templates, he opposes it on the view that “we should discuss the underlying issues” – that is, we should standardize the templates. Thus, he makes one argument in one case and exactly the opposite argument in another. All so that, in the end, he can engineer the outcome he wants: denying built-in pre-text in the etymology templates. There is another issue with the “consistency” line. Just as PUC believes uniformity should be “pursued for its own sake,” so too can naturalness. It is simply natural that a template that indicates borrowing and inheritance state so, especially when this would codify a preexisting norm.

The rest of the response lines up an elaborate plan to diminish as much as possible the usage of the templates, despite in the first instance agreeing that they advance a good cause. PUC argues that whenever the plus templates are used, “they could be replaced by the old templates by who[m]ever should wish to do so. They would be in a state of permanent semi-deprecation, if you want.” But at that point, “the templates might just as well not exist if they can never really be put to use.” This is analogous to the practice of switching, for example, label to lb en masse, without any meaningful changes to the edited entries themselves. We generally discourage this. But for the sole exception of the plus templates, PUC carves out a special and unprecedented edit-reversion license. Under it, any user can effectively ban the plus templates all on their own and undo the good edits of hard-working contributors. This regime is unheard of, and a similar suggestion for any of our other templates would be firmly rejected.

Reading the first line of PUC’s response, one might think it were a vote in support. PUC concedes that the templates are useful (at least for a subset of users). But he then immediately throws out a bevy of concerns intended to undermine the templates’ implementation. He suggests that the plus templates compete with prior ones, and bends over backwards to allow Victar the chance to undo other users’ edits. The facts remain, however, that the two sets of templates are noncompetitive, and that licenses to edit-war are not necessary. The only way to reach the opposite conclusions is by changing arguments in midstream and devising contrived means of reducing the templates’ reach.

"Let me be clear that I have absolutely no opinion on the matter of the bor+ and inh+ templates. However, it is very concerning that because a vote on the creation of templates has failed, some believe that the templates may not be created or used. Our convention is that failure of a vote represents a return to the status quo, which is (in this case) that the templates may still be created. A failed vote already is a nullity, and I support that concept. As I see it, the correct order of events would have been to (a) create the templates, (b) begin to use them in a few entries, then for someone to either (c) get approval for a bot to replace them into a large number of entries or (d) send them to RFDO. Although I support the principle of this vote (in fact I could have voted 'support per Llittleserie'), there would be some irony in me voting 'support' here, as I don't think this vote has any meaning either." (Abstain template removed; this argument was seconded by Kutchkutch, Ruakh, Geographyinitiative, and Guitarmankev1 ).

In response to nullification, abstainers speak with one resounding voice: The prior vote is already null and void, and there is no need to double down on that assertion. Abstainers share a great deal of common ground with supporters. Indeed, the great bulk of our reasoning is identical, up until the final stroke of logic.

The only divergence is abstainers’ argument that because the old vote “already is a nullity,” we should not state that explicitly to reaffirm the templates. As things were, the plus templates were in a state of constant limbo; and Mahagaja’s decision had unfairly advantaged opponents in striking down the usage of the templates. Without a declaration in favor of the templates, opponents could keep running victory laps and depriving everyone else the opportunity to employ them. This situation was untenable. The only solution was a univocal holding that the old vote could not be enforced and that templates do not need prior approval for creation. Nullification was our chance to make this be.

The template stipulation was a necessary evil
Ultimately, the nullification vote ended in no consensus, meaning that Mahagaja’s prior holding remained in force. If we did nothing else, the application of the plus templates would be in perpetual uncertainty, and supporters would always be at a disadvantage. Something had to be done to remedy this situation.

Examined in a vacuum, the best approach would have been to hold the line and insist that our position was correct. This would be the most principled approach. Supporters, the logic goes, should not yield to partisans on the other side, who themselves have refused to compromise on this issue. Why should supporters assume a defensive crouch and allow the hardline template-opponents to cheat their way to victory? According to this view, supporters must stand by the principles they fought for: The original vote was created on fraudulent grounds, it does not carry force to prohibit the templates, and if anything, it shed light on the overwhelming support for the plus templates. The best way forward was to continue advancing these ideas. In practice, this meant continuing to implement the plus templates despite the outcomes of the prior votes.

There is something to be said for this view. Being grounded in principles is a political virtue. Some users have tied themselves up in trying to make the opposite argument, a telling reflection on just how far template-opponents are willing to go to sustain their position. Perhaps when the opposition sinks to new lows in trying to save a failed argument, template-supporters should seize the high ground, doubling down on our commitment to principle. At bottom, we should not go along with an illegitimate vote. There was no reason to split the difference between the correct position (allowing the use of the plus templates) and an unreasonable one (prohibiting their deployment at every cost).

At the same time, the strategy of not taking no for an answer seems to get us nowhere. Contributors could use the templates in their edits, choosing to stand their ground against the past two votes, but Victar would have every advantage in undoing those edits. It would take us back to the proverbial “square one.” At some point, template-supporters should contend with the practical situations of the moment. The view that pure tenacity could have settled the matter was overly optimistic.

Properly framed, the question before us was not how to best force the templates through; but how to best secure a foothold to at least accomplish some of our aims. The stipulation compromised to do precisely this, but critically, it did not compromise on our values. Were it accepted, the stipulation would have put the bor+ back “in operation,” so that “anyone [would] be able to use it.” In addition, it would have safeguarded preexisting use of bor+ against systematic replacements. These are not peripheral matters; they are the most important points template-supporters argued for.

This is not to say that supporters should have accepted the stipulation and stopped there. After a few months, the next step would have been to propose mass bot replacements and finally standardize the templates. Those in favor of the inh+ template could have also asked to bring it back. But only after obtaining a limited victory could we have made any further progress.

In practice, the first approach seems to have won out among supporters, and it surprisingly worked. Victar did give up on trying to revert every use of the plus templates, although we are simply lucky that he did. He certainly did not have to, and if he decided to go back to his systematic template replacements today, he could do so with the blessing of Mahagaja’s ruling. So it is not time to rest on our laurels and drive this strategy forward in other matters of governance. This was a plan that happened to work in this particular case, but future cases should be handled with greater care and forethought.

We need an arbitration committee
The subtext of this essay is that we need an arbitration committee. The plus-templates situation has brought to light many of the deficiencies of our governance, and an arbitration committee is the most natural and effective way to supplement it. I will make more detailed arguments and proposals in a future essay, but for now I will preview three key arguments.

First, we need a body with the power of judicial review. Whoever sits on the board should be able to interpret our rules to invalidate violative votes. The entire controversy behind the templates situation would have been solved immediately if an arbitration committee had stepped in to clarify the rules. This body, by definition, would be a judicial branch regardless of whether we call it an “arbitration committee” or something else.

Second, a board formally endowed with adjudicative power is a better arbiter than any individual user alone could be. The board’s members would be elected and impeachable, meaning that they are doubly accountable to the will of the community. In addition, only users with sound judgment and discretion – as determined by our community – would hand down final verdicts. Furthermore, we would no longer be victim to the whims of individual administrators who, in some cases, wield virtually unreviewable authority. Again, this is a vital improvement to our current system, and one which would have avoided the arbitrary administration of the rules in the plus-templates dispute.

Third, an arbitration committee would be tightly bound by judicial conventions. It would review only live controversies brought to its attention by individual litigants. It would be tied down to bedrock legal principles of stare decisis, nulla poena sine lege, and res judicata among others. It would only be able to interpret our policies, rather than write them or enforce them. These are fixed restraints on the scope of judicial power, so that an arbitration committee would exercise “neither FORCE nor WILL, but merely judgment.”

* * * For the time being, the plus-templates situation has died down, and supporters have gotten much of what they wanted. But it is still worth analyzing the political decisions that were made while the templates were in question. Chief among them is Mahagaja’s unblock of Victar, which unfairly tipped the scales in favor of the minority and made a mess of the templates issue. It took a lot of work for the community to correct that error, including another no-consensus vote and a reconciliation effort to follow. Ultimately, we managed to restore the templates’ operation. Still, it is not time to rest on our laurels just because attacks on the templates have become less frequent. If they wanted to, any one template-opponent could effectively ban the templates under the aegis of Mahagaja’s holding. Even more troubling is not just how to deal with that potentiality, but how to resolve similar disputes in the future. At this point, it is hard to avoid the obvious solution: We need to establish an arbitration committee.