Wiktionary:Votes

Votes formalize and document the consensus-building process and the decisions that the community makes. This page displays the full contents of recent, current and planned votes.

See Wiktionary:Votes/Active to add new votes to the “active” list and remove old ones. Finished votes are added to Wiktionary:Votes/Timeline, an organized archive of previous votes and their results, sorted by the vote end date.

Creating a new vote

For the policy, see Wiktionary:Voting policy.

For a tutorial for creating a vote, see Help:Creating a vote.

Note: Update the vote list at WT:A.	Note: Update the vote list at WT:B.
Note: Update the vote list at WT:C.

Routine clean-up

For admins, periodically check for orphan votes at Wiktionary:Votes/ for votes and voting templates, including templates for creation of new votes:

• Pages starting with "Wiktionary:Votes".
• Pages starting with "Template:vote-".

Current and new votes

Planned, running and recent votes [edit this list]

(see also: timeline, policy)

Ends	Title	Status/Votes
Sep 27	Move constructed languages to own namespace	47 (27 people)
(=1)	Wiktionary:Table of votes	(=47)

Move constructed languages to own namespace

Proposal

Change 1

Create the Constructed namespace and move all Appendix-only constructed languages^[1] to the new namespace. Update setting $wgNamespacesToBeSearchedDefault^[2] to include the Constructed namespace.

Update the following WT:CONLANG passage from:

Other constructed languages should not be included as entries or translations in the main namespace, but may have lexicons in the Appendix namespace at the community's discretion. One use in a durably archived source is the minimum attestation required for an individual entry in an appendix-only constructed language.

To:

Other constructed languages should not be included as entries or translations in the main namespace, but may have entries in the Constructed namespace at the community's discretion. One use in a durably archived source is the minimum attestation required for an individual entry. All policies referring to the main namespace apply to the Constructed namespace as well, except where explicitly noted otherwise.

Change 2

If change 1 passes and the namespace is added to the default search criteria, update WT:CONLANG entirely to:

Constructed languages have not developed naturally, but are the product of conscious effort in the fulfillment of some purpose.

All constructed languages approved by community consensus shall be maintained in the Constructed namespace except for Eskayan. One use in a durably archived source is the minimum attestation required for an individual entry. All well-documented constructed languages may include translations under the applicable English terms in the main namespace. All policies referring to the main namespace apply to the Constructed namespace as well, except where explicitly noted otherwise.

Some individual terms from constructed languages have been adopted into other languages. These should be treated as terms in the adoptive language, and the origin noted in the etymology, regardless of whether the constructed language as a whole is included.

Background

Essentially, this vote is to maintain the current status quo regarding constructed languages, except for the following changes:

Change 1

• Move all appendix-only constructed languages out of the Appendix namespace and into its own namespace.
• Each page in the Constructed namespace may contain multiple languages, as is seen in the mainspace.
• Increase visibility of these languages by allowing them to be searchable via the search bar by default.

Change 2

• Move the remaining constructed languages from the mainspace to the Constructed namespace while maintaining the same visibility as if they were on the mainspace.

Eskayan is excluded from this vote because its language classification is uncertain across the community. That is, we do not know its status (rather than its status being contested). Currently, there is no formal discussion/consensus on how to handle translations of constructed languages and so, the vote's verbiage retains the current practice (allow WDL conlangs only). The setting-to-be-updated affects which results the user sees when they search for a term. For example, if they were to type amo, the following entries will be listed:

• amo
• Constructed:amo

This set of changes would increase accessibility to our coverage of constructed languages, as most of them are currently hidden away in the Appendix namespace. It is highly encouraged to read the linked discussion below, particularly starting at this comment. The discussion contains the reasons for both supporting and opposing in much greater detail.

Schedule:

• Vote starts: 00:00, 28 August 2026 (UTC)
• Vote ends: 23:59, 27 September 2026 (UTC)
• Vote created: TranqyPoo 💬 ✏️ 20:52, 18 August 2026 (UTC)

Discussion(s):

• Wiktionary:Beer parlour/2026/August § The issue of conlangs
• Wiktionary talk:Votes/2026-08/Move constructed languages to own namespace

Change 1

Support

1.  Support: this doesn’t by itself solve the many disagreements surrounding conlang entries (and it doesn’t intend to), but it will address one of the more important issues involved and improve the user experience by making conlang entries more accessible to readers (without mixing them with natural languages, which the community finds undesirable). — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 01:10, 29 August 2026 (UTC)
2.  Support: As the individual who initially proposed a separate conlang namespace in the context of this discussion, I unsurprisingly fully support it; it is thus regrettable that continuing health issues prevent me from justifying my decision. Hazarasp (parlement · werkis) 03:23, 29 August 2026 (UTC)
3.  Support make Appendix namespace less of a wastebasket. Lurker 320698738032 (talk) 11:16, 29 August 2026 (UTC)
4.  Support, provided bullet point #3 does happen. MedK1 (talk) 16:56, 29 August 2026 (UTC)

    Support as proposer. This is good middle-ground, but not my ideal set-up for constructed languages. My dream is that all CFI-compliant constructed languages will be featured in the mainspace, as they are just as important noteworthy as any language. To clarify: I do not advocate that all constructed languages should be covered in Wiktionary; only those that have demonstrated its worldly prominence. TranqyPoo 💬 ✏️ 17:12, 29 August 2026 (UTC)

5.  Support Vininn126 (talk) 17:35, 29 August 2026 (UTC)
6.  Support. I am not against the idea, although it still escapes me why any of this is necessary at all. The suggestion that moving languages out of mainspace will enhance visibility seems somewhat contradictory. Besides, I haven't heard a single convincing, meritorical argument for separating constructed languages from other languages, as if the former aren't really languages at all. I understand that the sheer number of (potential) constructed languages – varying from languages consisting of a few words only to languages with thousands of speakers – is kind of intimidating, but there are better, simpler methods to deal with that than a separate namespace. Besides, if separate namespaces for certain types of languages are really needed, why not do the same for ancient languages or pidgins, for example? Furthermore:
   • There is no dichotomic opposition between constructed and natural languages. Even the most natural language has constructed elements, and vice versa. Several languages belong to a grey area of language types that can easily be placed in both categories simultaneously.
   • I think it is a mistake to a create a namespace and move various languages there without taking an informed decision about criteria for inclusion first, such as: ISO 639 code, number of speakers, purpose, completeness, corpus, notability, durability, etc. As I already mentioned in the discussion, several of the languages currently in the Appendix are merely naming languages. Do we really want to give those more visibility?
   • As I have tried to point out, what matters to me personally are not so much the pages with words in constructed languages but rather their findability in the "Translations" sections of English words – or in a possible separate "Translations into constructed languages" section, for that matter. The way I understand it, if both parts of the proposal are adopted, there will still be two categories: languages with translations from English and languages without them. I think that's an undesirable situation. In my opinion, languages that aren't worth having translations from English shouldn't be here at all.
   • But that said, I can also see the potential advantages of the move, so I'll support the proposal, adding that it is my sincere hope that this new namespace is not going to be a dumping ground for stuff we want out of sight. IJzeren Jan (talk) 21:43, 29 August 2026 (UTC)

   A few remarks:

   • I agree that conlangs and natural languages are not fundamentally different types of entity, but disagree in contending the set of all languages can nevertheless be roughly bifurcated into the two groupings, each of which possess clear members (e.g. English, Klingon) and typical characteristics. The fact that some languages (e.g. N'Ko) are of unclear classification doesn't diminish the essential validity of this classification to me.
   • However, my aim isn't argue about conlang ontology; in fact, I suggested the idea of placing conlangs in a distinct namespace since I believe it is the only viable means of reconciling different editors' views regarding how conlangs should be handled on Wiktionary, many of which appear to stem from different conceptions of what conlangs fundamentally are and how they relate to natural languages:

   I believe creating a Conlang namespace is a solution worth considering; not only does this satisfy both conlang proponents' desire for more legitimacy than the Appendix namespace can afford and opponents' repulsion towards the idea of conlangs being treated equally to natural languages. it also gels nicely with the Reconstruction namespace. (In fact, a namespace for conlangs was previously suggested in the leadup to the Reconstruction namespace's creation.)

   • A potential reason why this vote doesn't alter conlang CFI or translation policy is to prevent it from getting bogged down in disputes regarding the proposals that'd be concocted for changing these, ensuring that it remains dedicated to assessing the fundamental merits of the conlang namespace concept. Hazarasp (parlement · werkis) 22:16, 29 August 2026 (UTC)

7.  Support I have taken part in the discussions and favor moving constructed languages to their own namespace; I have also mentioned being willing to support including translations into constructed languages in their own translations subsection (but not mixed in with mainspace languages). I am not a conlang fan in any way but I support the general trend here of avoiding privileging some constructed languages over others (beyond establishing minimum requirements of notability, completeness etc. that any conlang needs in order to be included at all). Benwing2 (talk) 01:37, 30 August 2026 (UTC)
8.  Weak support as a compromise per nom, with caveats listed per Polomo, Jan and (weakly) per AG. casting my vote, though unsure about what to say here other than repeat my points from BP. Juwan 🕊️🌈 09:08, 30 August 2026 (UTC)
9.  Support, hopefully this ultimately becomes a stepping stone to merging conlangs into mainspace. Whoop whoop pull up ^{♀️ Bitching Betty 🏳️‍⚧️ Averted crashes ⚧️} 01:41, 31 August 2026 (UTC)
10.  Support PhoenicianLetters (talk) 17:22, 31 August 2026 (UTC)
11.  Support Rcnn (talk) 16:17, 2 September 2026 (UTC)
12.  Support Rakso43243 (talk) 18:09, 3 September 2026 (UTC)
13.  Support Brewing ns Constructed: and moving all conlang terms that are already in ns Appendix: to Constructed:. Taylor 49 (talk) 13:48, 4 September 2026 (UTC)
14.  Support Tc14Hd (aka Marc) (talk) 15:23, 5 September 2026 (UTC)
15.  Weak support: I have the same view as Juwan. LunaEatsTuna (talk) 01:29, 11 September 2026 (UTC)

Oppose

1.  Weak oppose: I appreciate the time and effort that went into this vote; however, abefore the Construction namespace should be made default searchable and elevated, there needs to be a CFI, otherwise the current entry is small and any conlang published in one durably archived work would be included, which is less than what we currently require from Esperanto and the other conlangs currently in the mainspace (though Wiktionary talk:Votes/2026-08/Move constructed languages to own namespace § CFI indicates that there may be a conflict with wording). AG202 (talk) 03:15, 29 August 2026 (UTC)

   Just clarifying for other voters (this also applies to your vote under Change 2): the rules a language is subject to will not be different between the mainspace and the Constructed namespace, as specifically stated in the proposed policy text. The WT:WDL conlangs (currently in the mainspace) will continue being subject to WT:WDL criteria, and the WT:LDL conlangs (currently allowed in the appendix) will continue being subject to WT:LDL criteria. This vote is specifically designed to determine only where a languages’ entries go; it does not change any other part of applicable policy (e.g., how the entries look or whether they may exist). As such, this vote will not be the cause of many more conlangs being added to Wiktionary, because it does not touch our current standards for conlang entries. We talked more about this here. — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 03:33, 29 August 2026 (UTC)

   @Polomo: Yes, that doesn't dispute what I said though. The point was that the current policy does need to change if we're elevating the Construction namespace to be searchable. Several others mentioned this too in that thread. Our current CFI is very lax and allows things like Mandalorian ta'raysh, which can be cited in a durably archived source, but that I doubt is used outside of the Star Wars fictional universe. Do we want Eloi (which Wikipedia doesn't even have an article on) to be searchable? To be quite frank, I don't want those searchable because they are not as serious as the vast majority of our mainspace entries and should remain in our Appendix. We don't have the basic rules about a conlang needing to be cited outside of said fictional universe. AG202 (talk) 13:57, 29 August 2026 (UTC)

   I disagree. Fundamentally, this argument is based on the following question. Is it better to:

   • keep some languages hidden until there is a strict inclusion criteria, or
   • provide the same accessibility as the mainspace with the risk of readers observing languages that could be deleted?

   I side with the later. We should encourage growth in our coverage for all languages. I would prefer to lose already-recorded work due to a consensused decision than to hinder growth due us making it inaccessible. Imagine an inexperienced editor that is well-versed in a constructed language and whose activity of contributions only lasts a year. It is likely that they would not know about our coverage of their conlang-of-interest and therefore, we would lose the opportunity to increase our coverage of the constructed language. In other words, their unfamiliarity with our wiki (particularly, locating language categories and unsearchable-by-default namespaces) should not be a reason for a language's low amount of coverage. TranqyPoo 💬 ✏️ 16:26, 29 August 2026 (UTC)

   Imagine an inexperienced editor that is well-versed in a constructed language and whose activity of contributions only lasts a year. It is likely that they would not know about our coverage of their conlang-of-interest I discovered Category:Proto-Graeco-Phrygian_language by a complete accident soon after reading Šorgo 2026 and noticed that there were no entries. So this scenario is definitely possible. Lurker 320698738032 (talk) 18:30, 3 September 2026 (UTC)

2.  Oppose I'll be blunt, having spent so much time studying natural languages, and by doing so the history, thoughts and lives of the people that spoke them, I enjoy conlangs as much as a botanist would plastic flowers. I can live with Esperanto, it's got its own share of history, but the rest are good where they are. The agenda of this vote as explicitly stated on the Discord server, in particular the searchability of the namespace, is to gradually normalise conlangs into our minds to be on the same level as natural languages and eventually port them to the mainspace, and while I enjoy the egalitarian principle behind this view, I must disagree. One of the main factors for my participation to this project is language preservation, and floral biodiversity has nothing to do with plastic flowers. In this sanctuary of fact, I'm against giving prominence to delusions. Catonif (talk) 19:18, 29 August 2026 (UTC)

   Let's keep the derogatory language to a minimum please (i.e. delusions), so we can maintain civility and focus on the points. I'm sorry that you view this change as an agenda rather than an expansion to our inclusivity. Can you elaborate further on why constructed languages could not (or should not) be at the same notability level as natural languages? TranqyPoo 💬 ✏️ 19:57, 29 August 2026 (UTC)

   I apologise for the terminology, I was not aware that it has such a derogatory connotation, I would have used a more neutral term, like "imaginations", if I had. I do not mean anything inherently bad by "agenda" either. Rather than some twisted plot, I am mostly just referring to these two messages shared on 15 August by you on Discord. I think it will take time. The proposed upcoming changes will hopefully get editors used to the idea that conlangs are just as worthy of inclusion on Wiktionary as natlangs. […] Moving all constructed languages into their own namespace + search results including that namespace. I imagine some people may question "why are they separated when it is treated like it is in the mainspace". I hope not to be taken as a snitcher and not to undermine our friendship, it is precisely because I hold your rationanility highly that I am sure you will understand that this is only for fairness to those that choose not to participate in the Discord. And I also believe that a vote should be considered for what it changes, rather that the eventual ulterior motives of the vote creator, but in this case I am fairly convinced by your predictions, so I believe I am allowed to worry about it. I am all for inclusivity, and I am all for equity, which is why I like the principles behind arguments that many made in this movement, but there is always a line, unless you propose we document programming language as well. The presence of conlangs in our appendix is already more than I would have granted, a dedicated namespace I would have felt uneasy about but would have abstained, but the searchability is where I draw the line. Don't let my words worry you too much however, most probably the vote will pass anyways. I'm simply here to express my feelings. Catonif (talk) 22:24, 29 August 2026 (UTC)

   Esteemed colleague, I leave the terminology (delusions, imaginations) for what it is, but your comparison with plastic flowers is interesting, albeit not particularly flattering. You seem to forget one crucial thing here: that there are many different types of constructed languages. Some of them (like Klingon, Quenya or my own Boksirit) are simply a work of fiction, embedded in a larger work of fiction, and most people who study or even practice them are fans of the latter. Some (like Lojban or Toki Pona) are linguistic experiments, learned out of curiosity or just for fun. Some of them are scientific reconstructions of a hypothetical protolanguage. Some were meant to save the world with a common second language, learned by idealists and hobbyists. And some were meant to serve as an umbrella language for a group of related languages or dialects, perhaps even to mold them into one language with native speakers. I wouldn't compare any of those with plastic flowers. Plastic flowers are meant to look identical to real flowers. You can put them in a vase, but they won't grow if you water them, and they won't wither if you don't. A fictional language, however, is not at all intended to look like a faithful copy of a real language; if anything, you can compare it to a drawing of some fantastic, non-existing flower. Likewise, a reconstructed language is like a model of a flower that might have existed in the past. But when its comes to the last category, we are entering a domain that bears more similarity to something that actually does interest botanists: plant breeding and creating new races by mixing existing ones. Language of that type are more like cultivars, so to speak. IJzeren Jan (talk) 00:21, 30 August 2026 (UTC)

   You are correct: whatever I say in the Discord server is unobjectionably citable and votes should be transparent; I have no issue with this. My own hopes do not override community consensus; I accept that this vote may be the final resting place for constructed languages.

   To the matter at hand, you stated how you would treat conlangs, but did not provide further reasoning/justifications behind them. Why is the line drawn at natural languages? What makes conlangs worthy of segregation and hiddenness? I want to understand where your feelings are coming from, so that I may empathize and further explore our beliefs. TranqyPoo 💬 ✏️ 14:59, 30 August 2026 (UTC)

   @IJzeren Jan. Learned friend, it is a pleasure to make your acquantaince, unfortunate for this to be the setting. It is not my role or intention here to neither belittle nor flatter the existence of this phenomenon. The scope of the metaphor is to show that Wiktionary's search bar is simply not the place for these things, good or bad may they be. Though counterproductive, I'll keep the metaphor going, because I like them. As I see it, no matter the shape, colour or smell of the flower, or its inspiration, ideology or goal, or the innovativeness, historicity or unconventionality, the material is still inevitably plastic. On the other hand, despite how anthropic their conception is, cultivars still have biological ancestors, they are born, live and die as any other plant, and I would rather compare them to pidgins, or other kinds of lingue franche, which like the cultivars also have their mothers and fathers.

   @TranqyPoo. If I were to tell you that I swam from Sicily to Sardinia, you would likely be shocked, and rightfully so, as no other human has ever swum that kind of distance. If I told you, on the other hand, that rather than me it was a character from a book, your reaction would understandibly be quite different. Whether an event is real or not is not just one among many of its attributes, it is a crucial characteristic that goes as far as undermining its very identity as an "event".

   The reason why linguistics is such a relevant science in the first place is because it is a direct symptom of whatever molded it. From words and expressions you can reconstruct history, lifestyle and beliefs. From grammar and syntax your can study how our brain processes things, and how it can affect who we are. From the shape and colour of a flower you can know so much about the environment in which the plant evolved. The study of languages has two main goals, the first is communication, the second is this one, and it's what Wiktionary, the first omnilingual etymological dictionary humanity has ever worked on, truly excells at doing. The goal of studying conlangs differs from instance to instance, but it's never this, because there's no point in figuring out the distance someone swam if they never did.

   And finally, there's another reason, probably easier to understand, which is that these languages are still very niche compared to the various jargons of all the other fictional lexical bases. For each fluent speaker of Toki Pona there are a thousand people who played Plants vs. Zombies, and Harry Potter English, pass me the term, likely has a corpus large enough to overshadow some of our WDLs.

   I'm not here to give any qualitative reviews on these languages. If you don't care about the studying history of pollination, a plastic flower can last you much more and with much fewer water. I just hope to have shown that there is a crucial difference. Wikipedia does have articles on false events and fictional characters, but they need to be ten thousand times more notable than an existing equivalent for them to meet the inclusion criteria. Similarly, I do not oppose our inclusion of Esperanto, due to its impact, but there really isn't anything else quite like it. If you want to add all the side-characters of your favourite show, there's probably a Wikia for it, and if you want to document conlangs on wiki, we have the appendix, or even -- why not? -- its own dedicated namespace.

   But it should not be searchable, because as you said yourself, doing so is nearly equivalent to them being in the mainspace, and this they should not be. Catonif (talk) 18:47, 3 September 2026 (UTC)

   The plastic flower, swimming from Sicily to Sardinia and Wikia analogies are all quite illustrative of the difference between natural languages and constructed languages. In fact, I was thinking of mentioning the Wikia analogy in my own vote but eventually decided against it. It would be frightening if this were true:

   The agenda of this vote […] is to gradually normalise conlangs into our minds to be on the same level as natural languages and eventually port them to the mainspace.

   Kutchkutch (talk) 11:41, 4 September 2026 (UTC)

   As the individual who introduced the idea of a conlang namespace into this year's discussion, I am not party to this putative "agenda of the vote" and would be a vehement opponent of any attempt to introduce conlangs into the mainspace insofar as my health allows; this is not just because I concur with you that natural language possesses incomparable aesthetic qualities, but more importantly since it provides incomparable value to both linguistics and its sister sciences as a naïve, involuntary product of our linguistic powers.

   It must also be clarified that my idea was not intended to reflect any particular view regarding the nature of conlangs, their relation to natlangs, or how these should affect their treatment at Wiktionary, but to be an omnia omnibus that would function as a inoffensive ultimate settlement. However, I acknowledge that some indeed wish to "normalise conlangs" to various degrees; making the proposed namespace searchable is a harmless-seeming concession that I am willing to grant them, especially as the conspiratorial conlanger cabal seemingly consists of only some of this proposal's proponents.

   (In case you are unaware, Ido, Interlingua, and Volapük are also currently present in the mainspace). Hazarasp (parlement · werkis) 20:09, 29 August 2026 (UTC)

   You seem to have an agenda of your own (not that "having an agenda" is inherently bad), which is that constructed languages have no value and must thus be hidden away from users. You and other voters should consider whether their own dislike of constructed languages outweighs the usefulness that their proper documentation (and actual visibility) can have to our readers. I would expect many of our readers to be interested in conlangs and, among those, many to not know we document them as well. — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 20:43, 29 August 2026 (UTC)

   Besides the plastic flower analogy being clearly false, Wiktionary is not supposed to prescribe language use. It should not prescribe how constructed languages are to be treated by people in relation to natural ones either. Lurker 320698738032 (talk) 09:47, 30 August 2026 (UTC)

3.  Oppose I generally believe constructed languages have no place in Wiktionary. While I am indifferent to the namespace itself, I am concerned that establishing it will attract more artlangs and lead to lowered inclusion standards under the excuse that they are isolated in their own space. — Fenakhay ^{(حيطي · مساهماتي)} 19:00, 30 August 2026 (UTC)

   FWIW: AG202 and others are in the initial works to create a more defined (stricter) conlang CFI policy and so, I hope this would alleviate your artlang concern (and something I absolutely support). If you are willing to discuss, I would very much appreciate your elaborated opinion on why conlangs do not belong here. TranqyPoo 💬 ✏️ 20:27, 30 August 2026 (UTC)

4.  Oppose Provisions 1 and 2 are fine, provision 3 is not. I would not support default searchability for most of the currently appendix-only constructed languages either. — SURJECTION ^{/ T / C / L /} 19:11, 30 August 2026 (UTC)

   @Surjection, I agree with you in so far that IMO artlangs shouldn't be here at all – especially since several of the artlangs in question are naming languages (snippets of conlangs there were never even created). Artlangs are works of fiction, rarely intended to be spoken by anybody at all. That's what I meant when I wrote that we should have discussed the CFI before moving stuff to new namespaces. What I don't understand, however, why searchability should be a problem. Having languages here but insisting that they shouldn't be searchable is like saying: "You are welcome to join our club, but only if you make sure that nobody can ever see or hear you." IJzeren Jan (talk) 10:10, 31 August 2026 (UTC)

5.  Oppose Wiktionary should prioritise the coverage of natural languages. There are countless understudied and vulnerable natural languages that could benefit from more attention. As per Catonif, natural languages represent the history, thoughts and lives of the people that spoke them. I concur with Fenakhay that the namespace would likely attract more artlangs. I also agree with Surjection on I would not support default searchability for most of the currently appendix-only constructed languages. Kutchkutch (talk) 05:25, 31 August 2026 (UTC)

   My compliments, @Kutchkutch, you just gave a beautiful example of a false dilemma. How do you think banning constructed languages will help vulnerable natural languages? Do you seriously believe that those who work with constructed languages will massively start travelling to Lesotho or Papua New Guinea to study some dying tribal language instead? It's not like I haven't seen this argumentation before, but honestly, it is like saying that football games should be forbidden because people who watch them should do something more useful instead. For the record, many conlangers are professional linguists, and many conlangers actually do study vulnerable languages. IJzeren Jan (talk) 10:32, 31 August 2026 (UTC)

   • As AG202 alluded to, being exempt from WT:Criteria_for_inclusion/Fictional_universes is probably why many of these conlangs still exist in the appendix and haven’t been deleted.
   • Increasing coverage of, maintaining and building infrastructure around less serious conlangs entails less time spent on natural languages, since time is a limited resource.
   • In addition, if a reader doesn’t notice the Constructed namespace and hasn’t ever heard of a particular conlang, then it may look like a natural language. For example, the name of the conlang Toki Pona could easily be confused with the creole Tok Pisin. At least reconstructed terms have a preceding asterisk and the language name may have the prefix proto-.
   • Wiktionary certainly isn’t asking its editors to travel long distances to document natural languages. There might be enough information to document languages that interest editors online. In any case, if this change passes, then may the Force be with you for documenting conlangs. Kutchkutch (talk) 12:23, 2 September 2026 (UTC)

     2. Wiktionary is edited by volunteers who work on whatever they want. Time spent on any language depends on editors' attitude, not on existing coverage.

     3. There are also similarly named natural languages, e.g. Slovak and Slovene or Frisian, Friulian and Faroese. Toki Pona being named similar to Tok Pisin does not increase ambiguity significantly. Lurker 320698738032 (talk) 13:38, 2 September 2026 (UTC)

     Indeed, the Slavey language might be mistaken for a Slavic language, East Frisian is a Low Saxon and not a Frisian dialect, and there are unrelated languages sharing the same name (Ket, Soyot). But is that really so much of a problem? Languages are named as they are, no matter if they are constructed or not. And indeed, there is no guarantee that people working here on constructed languages would otherwise be working on something else.

     But @Kutchkutch is right about fictional languages, and that is precisely why I have been arguing that we should worry about criteria first before creating new namespaces. Of the 27 languages currently in the Appendix, at least nine are way too incomplete to be considered (full) languages at all (as you can see in a table I made on the talk page). Another thing is that artistic languages in general aren't even meant to be spoken by anybody: there might be a community of fans who enjoy them, but it's not like they are used in daily life. And one last thing: artistic languages are the intellectual property of their creators (or other copyright holders). Languages as such cannot be copyrighted, but dictionaries can. And being such a creator myself, I can assure you that I really wouldn't appreciate it if others started making dictionaries of my languages – let alone copying my dictionaries to spaces I cannot control. IJzeren Jan (talk) 21:58, 2 September 2026 (UTC)

     at least nine are way too incomplete to be considered (full) languages at all Are they one-page appendices? Then this will not affect them much. Lurker 320698738032 (talk) 13:37, 3 September 2026 (UTC)

     • The fictional AP:Klingon and AP:Quenya languages have over 2,000 entries. It would be unfortunate to see those entries be deleted because of the WT:CFI.
     • It would have been better if only certain conlangs such as AP:Interlingue and CAT:Lojban language were to be moved to the Constructed namespace rather than moving all of them. Perhaps the single-page appendices at T:Artistic languages should stay in the appendix namespace.

     Kutchkutch (talk) 15:32, 3 September 2026 (UTC)

     I looked at the subpages with Klingon entries. Considering what I was told at Wiktionary:Beer_parlour/2026/August#Converting_PGPh._pages_to_appendix_pages, these should be collapsed into a one-page appendix or possibly one page per part of speech. Lurker 320698738032 (talk) 18:41, 3 September 2026 (UTC)

     Perhaps the single-page appendices at T:Artistic languages should stay in the appendix namespace I thought that was the point. Moving entry pages to Constructed does not mean that all constructed languages must have entries there. Lurker 320698738032 (talk) 18:45, 3 September 2026 (UTC)

     @Lurker 320698738032: The vote language explicitly says Other constructed languages should not be included as entries or translations in the main namespace, but may have entries in the Constructed namespace at the community's discretion. One use in a durably archived source is the minimum attestation required for an individual entry. As long as a Constructed language falls within those criteria, it can be included in the Constructed namespace, and all Appendix-only conlangs already pass those rules. As such, they will all be moved to the new namespace if this vote passes. For example, an incomplete artificial language with no speakers, Mandalorian, which doesn't even have a Wikipedia page, has entries here like ta'raysh, which does have a use in running text. There's no strict CFI, and what you were told at the PGPh discussion is unfortunately not as relevant to this discussion. AG202 (talk) 13:01, 4 September 2026 (UTC)

    Oppose because i dont see anyone addressing the extremely low CFI barrier of just one durably archived use yet. The Usenet loophole is still wide open, we've never decisively ruled out self-published media, and our social media policy is ambiguous but we've been declaring social media posts durably archived either directly or through a third-party archival service. Besides the fact that this could lead to people uploading their conlangs' entire lexicons with little effort, there's also a copyright issue, in that some conlang authors might not want their work copied onto Wiktionary. I'm open to changing my vote since my rationale is different than everyone else's so far and although I don't think we can change the text of the proposal in the middle of the vote, I would be willing to accept assurance that we'll have some way to prevent what I'm worried about. I can name other examples of things that we could be doing but which, as a community, we seem to have collectively discouraged it without a written policy against it. —Soap— 16:18, 3 September 2026 (UTC)

   @Soap: Please review my comment. I don't think it would sway you, but it can help you gain a perspective on why one would support this with your concern in mind. I encourage you to stick to your guns, as this vote can be re-proposed after the CFI concern is addressed. Additionally AFAIK, people cannot add languages (even constructed) without consensus (through WT:LTR). TranqyPoo 💬 ✏️ 16:34, 3 September 2026 (UTC)

   I appreciate your prompt reply. Thank you. So, if I understand right, youre saying the two conditions

   but may have entries in the Constructed namespace at the community's discretion.

   and

   One use in a durably archived source is the minimum attestation required for an individual entry.

   are independent? so even if i have 1000 words in a print book of Conlang X, all of which are uses in running text rather than mentions (just to test the strongest possible case), we must first have a vote to allow the language to be entered at all? I had assumed that because I'd never seen a discussion about allowing any particular language into the Appendix namespace, that the community discretion criterion was a formality that we might never actually use, or that the act of entering an entry was all the consensus we needed unless that person was immediately reverted. Perhaps I've missed the discussions about allowing each specific language into the Appendix namespace, or maybe the ones we have there were added to the Appendix before that criterion was added to the policy. In either case, I just want to understand the new policy better. Will we require a formal discussion before allowing each new language into the Constructed namespace? Thanks, —Soap— 16:48, 3 September 2026 (UTC)

   @Soap: I believe your understanding is correct. To answer your question at the top, the first determines whether a language is approved to be included and the other is for individual entries of approved languages. That is, if a language is unapproved, the latter condition is irrelevant as all entries in that language would be deleted regardless.

   There have been courteous/curious discussions for adding some conlangs, like Wiktionary:Beer parlour/2026/April § Lingua Ignota and other conlang entries. However, I don't believe there is a formal written-in-stone policy yet to do so. The current prevention measure is that a language must have a language code. Those cannot be created without either having an ISO code (hard to achieve) or one of our admins making one (which requires a discussion in WT:LTR). If this measure does not satisfy you, I think it would be best that you continue to oppose.

   I can assure you (but not promise) that the community is motivated and seeking to make a CFI policy specifically for conlangs. After this vote, I intend to round up the key players to draft a policy and present it to the community for critique, then a vote. I anticipate that, upon implementing the new CFI policy, a handful of conlangs will be deleted. TranqyPoo 💬 ✏️ 17:41, 3 September 2026 (UTC)

   i would like to change my vote to neutral. i dont know whether it's better to move this whole discussion to the abstain section or to indent it and leave it in place. i'm worried that merely indenting it will lead people to still read it as an oppose and also that the abstain section is intended for simple votes rather than long discussions. what should i do? i dont feel comfortable changing to support because there are too many open questions. —Soap— 18:56, 3 September 2026 (UTC)

   @Soap: I recommend indenting your vote comment and striking the vote template, then creating an abstain vote. I believe it is important for people to follow your train of thought (leading to your abstain vote). You can point to this discussion in your abstain vote if you'd like, but I think the readers would be able to connect the dots. TranqyPoo 💬 ✏️ 19:50, 3 September 2026 (UTC)

6.  Weak oppose I fully support points 1 & 2, though I am honestly conflicted as to whether point 3 is a good idea... — BABR・kurwa? 18:59, 3 September 2026 (UTC)
7.  Oppose. AG202 made a good point in their comment. Specifically, single-page conlang appendixes will have to be moved as well based on this vote's verbiage. This would put much unnecessary strain on the editor(s) who will be moving these languages when the languages are likely to be deleted anyway after we vote in a new conlang CFI policy. The strain would stem from creating individual entries when they are all located on a single page (not likely botable). I would change my vote if the person who would be moving these terms can confirm that it would not be much effort to move these and later delete them. TranqyPoo 💬 ✏️ 13:38, 4 September 2026 (UTC)

   @TranqyPoo What makes you think that the entries would be moved by humans, and not bots? — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 16:48, 4 September 2026 (UTC)

   @Polomo: Compare: Quenya; it seems like it would take a lot of manual effort to set up a bot job. Particularly, it's untemplatized headwords and non-header See also and References sections. Disclaimer: I may have misplaced my concern in my Oppose comment. My concern is more: is the juice worth the squeeze to move, then delete them? I am not one who bots. It may seem easy, but until I know that it is, I cannot support this vote in good faith.

   Additionally, opposing this vote isn't a bad thing. This measure can be revoted with the people's concerns addressed (and therefore, more people who feel comfortable with this idea; one that everyone would be willing to back and defend, instead of feeling like they weren't heard. You can read more about my approach here). A vote that passes by a narrow margin does not make me feel comfortable. There is no rush to implementing this solution. We have waited years for a change, what is another few months? TranqyPoo 💬 ✏️ 18:24, 4 September 2026 (UTC)

   While I still vote in favor of both changes, I think that constructed languages with stub entries (i.e. only IPA and one sense) should be converted to one-page appendices (or per-PoS pages but not per-entry) instead of moving those entries to Constructed. Lurker 320698738032 (talk) 10:39, 5 September 2026 (UTC)

8.  Oppose I think it's important a distinction is made between auxlangs with established communities such as Esperanto and the aforementioned Eskayan, and artlangs such as Quenya that aren't intended for communication, and I think the current system fulfills such a distinction sufficiently. Whether certain languages need to be moved around namespaces might be a separate discussion worth having; for instance, I think Toki Pona might have a case for being brought to mainspace. Moreover, another user has pointed out that Nynorsk or Modern Hebrew could technically be considered constructed, and one could even argue at a stretch that some sign languages are constructed, so I think the distinction to be made is not by their origin but by their use. Tesco250 (talk) 14:14, 4 September 2026 (UTC)

   @Tesco250: But change 1 would not affect this distinction between mainspace and appendixspace conlangs; it would only make the latter more searchable. jlwoodwa (talk) 22:52, 5 September 2026 (UTC)

9.  Oppose tl;dr per Tesco250. Full arguments:

   If I want to know about Esperanto as it was intended by Zamenhof, I’ll read Zamenhof. And if I want to know about the language some random dude on the internet made up yesterday, I’ll ask the random dude. That is not what Wiktionary is about. Now, if I want to know about Esperanto as it is actually used by people to communicate with one another, that’s a very different story. Actual Esperanto is not the same as what Zamenhof wrote about, and that is something Wiktionary should (and does) document. But if a random dude made up a language yesterday, it won’t have a speech community yet, and it certainly won’t have any use that deviates from the random dude’s design. So nothing for Wiktionary to do there really.

   In my opinion, the only distinction that should matter for us is whether a language (conlang or otherwise) is more like Esperanto (an actual speech community has been using it for their communication, and in the process the language has been evolving with their needs) or more like a language some random dude on the internet made up yesterday (no speech community, just the random dude’s specs). I really don’t see the point of any of the other arguments that have been made (and usually debunked) so far.

   If a language is in use, we should put it in namespace. The conlang–natlang divide is just as artificial as any conlang, as pointed out by others several times. And if it is not in use, just remove it. That’s what we have RfV for. If a language is in use, and most of this use was durably archived (as is usually the case with conlangs, but not so with natlangs), then three independent citations will show how its words are in use. If a language is not in use, I really don’t see why it should be allowed in an appendix or a “constructed” namespace (searchable at that), while the natlang words I made up yesterday are mercilessly deleted.

   – MuDavid 栘𩿠 (talk) 02:36, 11 September 2026 (UTC)

   Quantifying this distinction is the subject of arguments above. It is also not clear from your comment what you propose instead. Lurker 320698738032 (talk) 14:11, 11 September 2026 (UTC)

   See my last paragraph. MuDavid 栘𩿠 (talk) 02:13, 12 September 2026 (UTC)

10.  Oppose The new CFI should have come first. And bullet 3 should have been split from bullet 1 and bullet 2. Then I would support all the changes. Zbutie3.14 (talk) 15:16, 12 September 2026 (UTC)

Abstain

1.  Abstain per this comment and the following discussion. —Soap— 21:53, 3 September 2026 (UTC)

Change 2

Support

1.  Support. Supporters of option 1 should support this as well, because the alternative is poor user experience – entries for conlangs would be split across two different namespaces, and users would be confused in finding an entry for a conlang outside of the dedicated “Constructed” namespace. The original reason for placing Esperanto etc. entries in the mainspace was for them to be easily accessible, and this change will maintain their accessibility. Note that the updated policy verbiage is mostly done to reflect what is already the status quo; it is not intended to reflect a complete and final decision in our handling of conlangs (that would be the product of future votes). — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 03:16, 29 August 2026 (UTC)
2.  Support: While I will reiterate that I am currently in no state to fully justify the move of WDL conlangs into the mainspace, I feel a few points are nevertheless worth reiterating:
   • As Polomo notes, keeping languages typically classified as conlangs outside of the appropriate namespace will likely engender continual confusion.
   • The origin/provenience of a language, as opposed to its current status, is both the most intuitive and the only ultimately indefeasible principle for dividing languages into conlangs and natural languages; this applies to namespace placement just as much as to the terminology.
   • Regardless, native speakers are a uninfluential minority within all conlang communities that reportedly possess them, so any supposition that they somehow imbue certain conlangs with a distinct character or any claim that their linguistic practice is authoritative is spurious.
   • I cannot agree with the claim that revived or standardised languages can be conceived of as conlangs since the modification or redaction of a extant body of linguistic material is a distinct process to the invention or synthesis of a novel one, regardless of possible ambiguity in certain instances (e.g. N'Ko).

   Hazarasp (parlement · werkis) 02:48, 29 August 2026 (UTC)

   @Hazarasp: I'll only respond to point 3: Esperanto is the only constructed language independently verified to have native speakers. Fiedler 2011 makes that clear, and it was cited in the main thread. The same source stated that while the native speakers do not have that much authority (likely due to other reasons), their speech has phrases, constructions, onomatopoeia, and more that are unique to them. That in and of itself means that they are developing natural language, and there've yet to be updated studies on the now third-generation natives. Other constructed languages cannot say the same. That, along with the other reasons listed in my oppose vote, at the very least gives one major reason to make the principle that the "origin/provenience of a language" being the way we handle languages not "indefeasible". (Let alone how that affects certain natural languages). AG202 (talk) 03:26, 29 August 2026 (UTC)

   My claim that origin is/should be the pertinent criterion when demarcating between conlangs and natlangs shouldn't be interpreted as a statement that origin as a conlang means a language isn't amenable to developing features stereotypically associated with native speakers such as generational transmission (and/or that conversely, natural language origin prevents constructed features); it instead constitutes a assertion that such features are not necessarily dispositive. In fact, it is not difficult to argue that most or even all conlangs possess "natural" elements and vice versa. Hazarasp (parlement · werkis) 04:11, 29 August 2026 (UTC)

3.  Support As I said earlier, Esperanto should not be treated differently from other constructed languages. Contra @AG202:
   • Tracking in Ethnologue does not matter (does it track Sanskrit, which is also arguably constructed?);
   • Native speakers do not make Esperanto a non-constructed language;
   • Lojban is not dependent on creator's whims either;
   • Same as #2. Esperanto really is a constructed language.

   Lurker 320698738032 (talk) 11:32, 29 August 2026 (UTC)

   @Lurker 320698738032: 1. Yes, yes it does. That took two seconds to verify. 2. Never said that it's not, was very explicit about that. 3. Wikipedia seems to disagree with you, also said creators' plural. 4. same as 2 AG202 (talk) 14:01, 29 August 2026 (UTC)

   Even Wikipedia admits that The speakers of Lojban are now free to construct new words and idioms, and decide where the language is heading. On top of that, Lojban use (at least on the internet) employs a lot of "unofficial" extensions. I am certain that there are other similarly uncontrolled constructed languages. Lurker 320698738032 (talk) 15:57, 29 August 2026 (UTC)

   That does not mean that it doesn’t rely on the creators’ whims. And with the cited very low number of speakers, it’s not at the same level as Esperanto, but that’s neither here nor there. AG202 (talk) 17:27, 29 August 2026 (UTC)

   Lojban uses a centralized dictionary where anyone can create a word. This dictionary is crucial to understanding Lojban; without it, a speaker would find conversations (particularly, the exchange of ideas) to be quite difficult.

   @AG202, I must ask: what experience do you have with constructed languages? I notice that you tend to use references to support your claims, but your claims do not match how the average speaker understands the languages-in-question. I'm getting "scientists state the sun does not rise or set, so we shouldn't say sunrise and sunset" vibes. TranqyPoo 💬 ✏️ 16:49, 29 August 2026 (UTC)

   I find that it’s much easier to back claims up with direct evidence rather than anecdotal evidence. Otherwise we’ll keep making decisions based upon faulty evidence, as has been done or almost done in the past. AG202 (talk) 17:26, 29 August 2026 (UTC)

   I must kindly be direct: please answer my question about your experience. As a person who can be easily swayed in arguments, I find this aspect important. TranqyPoo 💬 ✏️ 17:36, 29 August 2026 (UTC)

   I have studied Esperanto in the past, have read through many studies related to and grammars of constructed languages, and have been "lurking" in constructed language spaces for over a decade. However, I would not say that I've been active in any conlang space or that I speak any constructed language (as my infoboxes show). I've been more of an observer than anything. AG202 (talk) 18:28, 29 August 2026 (UTC)

   As far as I know, there are multiple Lojban dictionaries (including Wiktionary). Like with natural languages, if one cannot find something in one dictionary, one would see another.

   I would not call Lojban centralized. These words may introduce new syntactical features, which is rare in Esperanto. So in one manner Lojban is less centralized than Esperanto. Lurker 320698738032 (talk) 15:53, 31 August 2026 (UTC)

   I respectfully disagree, but I think this particular topic has no effect on the vote. Feel free to include me in a discussion elsewhere if you want to explore further. TranqyPoo 💬 ✏️ 16:25, 31 August 2026 (UTC)

4.  Strong support. Esperanto is the poster child for constructed languages, and I'm not very convinced by AG's arguments. If the fact that it has native speakers means it should be subject to the rules of natural languages, so be it: let's treat it as a natural language in the Constructed namespace. MedK1 (talk) 17:00, 29 August 2026 (UTC)

    Strong support Without this change, there will be an unintuitive division between constructed languages listed in the mainspace and the Constructed namespace. "Is the language constructed?" should be the only consideration. TranqyPoo 💬 ✏️ 17:18, 29 August 2026 (UTC)

5.  Support Vininn126 (talk) 17:35, 29 August 2026 (UTC)
6.  Support. I agree that supporting change 1 should automatically entail supporting change 2 as well, so there's that. And this should absolutely include Esperanto as well. Not only because excluding the only constructed language that everybody knows as such – the flagbearer of all constructed languages, so to speak – from the process would be hard to explain, but also because doing so would automatically reduce all other constructed languages to something futile. Like it or not, interlinguistics is an estabilished scientific discipline that studies a lot more than just Esperanto. Now, I am absolutely not saying that native speakers are meaningless, but it remains a fact that Esperanto was never even meant to have native speakers, since the objective for it was to become a universal second language. Measuring the importance of an international auxiliary languages by its number of native speakers is like measuring the importance of a book by the number of films based on it. Besides, it is nowhere said that a constructed language cannot have native speakers. Also, native speakers are by no means a condition sine qua non for natural development for a language. As far as Esperanto has undergone any changes since its creation by Zamenhof, they were mostly induced by L2 users, which shouldn't come as a surprise, since native speakers constitute a tiny minority within the community. So why shouldn't the same work for other languages with large user communities as well? IJzeren Jan (talk) 23:27, 29 August 2026 (UTC)
7.  Support Benwing2 (talk) 01:37, 30 August 2026 (UTC)
8.  Support PhoenicianLetters (talk) 17:39, 31 August 2026 (UTC)
9.  Support Rcnn (talk) 16:18, 2 September 2026 (UTC)
10.  Support Rakso43243 (talk) 18:10, 3 September 2026 (UTC)

Oppose

1.  Strong oppose: I've gone into this in significant detail at Wiktionary:Beer parlour/2026/August § The issue of conlangs and previous discussions (see: Wiktionary:Votes/2024-06/CFI for mainspace constructed languages and the linked discussions there), but to boil it down to my main concern: Esperanto should be excluded from the Construction namespace and the rules that govern it (especially without the necessary changes to CFI). Though it is a constructed language, in practical effect and in other locations, Esperanto is much closer to a natural language than the rest of the constructed languages, so we should treat it as one: 1) It is the only constructed language that Ethnologue tracks. 2) It has a critical mass of verified native speakers, up to the third generation, with it being a primarily spoken language for them. 3) Per Lindstedt in his 2011 paper, Esperanto has gone through several persistent changes over the years through its speakers, so it doesn't solely rely on the creators' rules and whims. Its attestation rules and its corpus are completely different from the rest. 4) It has a strong and documented literary and oral history among verifiably fluent and native speakers. I acknowledge the possible concerns about Esperanto not being in the Constructed namespace, but I strongly believe that it should be listed amongst natural languages and be subject to those rules, for comparison and scholarship's sake. Though this may be moot in the end since we don't yet know if the searchability aspect would be implemented. AG202 (talk) 03:15, 29 August 2026 (UTC)
2.  Oppose Again, I am to believe this change has the same agenda I described above. Bringing Esperanto down to the level of other conlangs, so that people will feel uneasy about Esperanto being separated, and as a solution all will be moved to the namespace. Esperanto needs to stay here, the other ones can go. Catonif (talk) 19:26, 29 August 2026 (UTC)
3.  Oppose moving Esperanto out of mainspace. It seems to me that this would be very disruptive to Esperanto for no reason other than consistency (which is an important but not entirely overriding consideration). If we were to consider Esperanto in its own rights, a vote to remove it from mainspace would almost certainly fail; I don't think it's right for it to be swept up along with everything else. I would be willing to re-evaluate this if and when other languages were moved individually to this new namespace and we could see the aftermath of such a move. jlwoodwa (talk) 01:40, 30 August 2026 (UTC)

   What do you consider "disruptive"? Just the process of moving the entries? — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 04:16, 30 August 2026 (UTC)

   Please elaborate on the disruption you mentioned. From my perspective, these changes would not affect Esperanto's visibility or maintainence, but you may have a better view than I. TranqyPoo 💬 ✏️ 04:53, 30 August 2026 (UTC)

   Not only do I agree that a hypothetical proposal that only consisted of removing Esperanto from the mainspace would probably garner insufficient support, but I would oppose such a proposal myself. However, that's rather besides the point; divorcing a individual stipulation from the overarching proposal gives a mistaken impression of both individuity and the entirety, thus tarnishing any evaluation of them whether in the context of this vote or elsewhere. Hazarasp (parlement · werkis) 10:58, 30 August 2026 (UTC)

   BTW for other folks, jlwoodwa is, AFAICT, one of the most active Esperanto editors on this project, if not the most active one. So I trust their judgment from the Esperanto editor angle. AG202 (talk) 13:16, 30 August 2026 (UTC)

   @Polomo, TranqyPoo: Here's what comes to mind for me:

   • Links of the form [[amo#Esperanto]] will be broken.
     • Some, but not all of them, can be retargeted.
       • On this wiki, links generated by templates will be easy to retarget, and manual links can be fixed by a bot.
       • Links on other Wikimedia wikis, whether through templates like w:Template:wikt-lang or manual links, will be harder to find and correct; this might take a long time.
       • Links from other websites generally cannot be fixed.
     • As long as we have hatnotes (like {{also}}) at the top of relevant mainspace pages, linking to the Constructed page with the same title, the harm of these broken links is reduced (if not eliminated).
       • But will we have these hatnotes? Some people will probably oppose them. I'm not sure what the general opinion would be.
   • There are many scripts, tools, and other software that only work in mainspace.
     • Some, like WT:ACCEL, will probably be fixed before too long.
     • Others, like d:User:Nikki/LexemeInterwikiLinks.js, might take a while to be fixed.
     • The "Cognate" MediaWiki extension, which provides links between each language edition of Wiktionary, explicitly states that it does not support non-standard namespaces.
   • There are unresolved design/policy issues.
     • For instance, how will the Citations namespace work? There are many Esperanto citations in that namespace already. If they are allowed to stay, will all the other constructed languages be allowed there too?
   • There are also the general costs (not specific to this discussion) of any move like this; I wouldn't oppose for these alone, but they're still a factor.
     • Many entry histories will be split between pages, making it harder to navigate those histories.
     • Until Google indexes the new pages, googling an Esperanto word will either give a broken link or no English Wiktionary results at all.

   And the main downside of not moving – reader confusion – can be addressed in other ways, so it's alright if we take our time here:

   • If someone enters the term "enabismiĝi" into the Wiktionary search bar, and they get a single result, they'll click on it and find what they were looking for, regardless of whether that result's title is "enabismiĝi" or "Constructed:enabismiĝi". We don't need to worry about this.
   • For the cases where an Esperanto word is spelled the same as a word in another constructed language (like Esperanto mi and Toki Pona mi), a dummy entry at "Constructed:mi#Esperanto" that simply directs readers to "mi#Esperanto" would only be slightly less convenient for those readers, while avoiding all the issues with actually moving the entry. These dummy entries could be easily created by a bot, and we wouldn't need all that many of them. I ran a few scripts, and found that:
     • If we moved languages from the appendix to the Constructed namespace, and didn't move any from mainspace, we would need 1,304 dummy entries in 998 Constructed pages.
     • If we moved Volapük as well, we would need 1,096 dummy entries in 846 pages; if we also moved Interlingua, we would need 1,482 dummy entries in 1,003 pages.
     • If we moved Ido too, leaving only Esperanto, we would need 5,300 dummy entries. (The larger number here is due to Ido being derived from Esperanto.) This suggests that it might be best for Esperanto and Ido to move together. Also supporting this is the fact that there are 2,979 pages with only Esperanto and Ido entries; if they're moved simultaneously, the page itself could be moved, which is more convenient and avoids splitting page history.

   With all this in mind, I don't think many people would object to leaving Esperanto as one of the last languages to be moved. But I'm going one step further and opposing for now, because I don't think we should decide whether to move Esperanto until we know more. Of the issues I listed above (and the ones that we won't know about until they happen!), some will be fixed quickly, and others could take years or just never be fixed. But how many will be in that second category, and how much of an impact will they have? Until we know that, I can't support moving Esperanto. jlwoodwa (talk) 23:15, 31 August 2026 (UTC)

   @jlwoodwa: Thank you for providing your perspective; none of your points crossed my mind!

   • How will the Citations namespace work?
     • I assumed that it would work identically to how we treat the Appendix fictional terms. Namely, they all share the Citations namespace. We currently do this with fictional words, like Citations:mutant and Citations:peanut. Although admittedly, this practice is rarely used.
   • Regarding dummy entries:
     • I very much like this idea! However, I'm a little confused. If we do not move languages from the mainspace, wouldn't there be more dummy entries in the Constructed namespace as there would be more languages to be soft-redirected? How your data is presented seems like there would be less. And to clarify, a bot would need run this job routinely as entries get created over time. Either way, this is enough for me to change my vote.

   TranqyPoo 💬 ✏️ 02:47, 2 September 2026 (UTC)

   Disregard my confusion! There would be less dummy entries because there would be less constructed language pages in the Constructed namespace due to less languages in the namespace that share spelling with the mainspace constructed languages. TranqyPoo 💬 ✏️ 03:40, 2 September 2026 (UTC)

4.  Oppose Notable constructed languages that have acquired native speakers should remain in the mainspace. — Fenakhay ^{(حيطي · مساهماتي)} 19:00, 30 August 2026 (UTC)
5.  Oppose as someone who believes that conlangs should generally be in mainspace. Whoop whoop pull up ^{♀️ Bitching Betty 🏳️‍⚧️ Averted crashes ⚧️} 01:42, 31 August 2026 (UTC)

   Understand that, by supporting Change 1 and opposing Change 2, constructed languages will be split across two namespaces. When a user wants to look up amo, they will be confronted with two options: amo and Constructed:amo. Since Esperanto is a constructed language, the instinctual choice would be Constructed:amo, but Esperanto's entry will not be there.

   If the issue is "Esperanto belongs in the mainspace", this vote effectively makes the two namespaces identical in both formatting and accessibility. Is there some disadvantage I am not seeing? TranqyPoo 💬 ✏️ 04:31, 31 August 2026 (UTC)

   @Whoop whoop pull up, pinging in case you would not see the above message otherwise. It is a very important consideration IMO. — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 15:20, 31 August 2026 (UTC)

   Conlangs already _are_ split across two namespaces (main and Appendix). Change 1 at least gets the Appendix ones out of the current junkbasket namespace into a new one, and, thus, is at least mildly helpful (if nothing else, it disassociates conlangs from other less-mainspace-worthy inhabitants of the Appendix), whereas Change 2 would be a retrograde step by removing from mainspace the conlangs that are currently there.

   Of course, in my view, a better choice would be to have all conlangs in mainspace, but that's not one of the options being offered here. Whoop whoop pull up ^{♀️ Bitching Betty 🏳️‍⚧️ Averted crashes ⚧️} 17:53, 31 August 2026 (UTC)

   @TranqyPoo Whoop whoop pull up ^{♀️ Bitching Betty 🏳️‍⚧️ Averted crashes ⚧️} 17:54, 31 August 2026 (UTC)

   @Whoop whoop pull up Mainspace conlangs will remain as accessible as they are currently, because the "Constructed:" namespace will also be accessible from the searchbox. The vast majority of our users utilize Wiktionary by searching for specific words whose meanings they want to know, and this will be as easy to do as it is now. There is nothing inherently better or privileged about being in the mainspace (same policies and accessibility), unless you think that being listed next to English and Estonian is a necessity. Meanwhile, there is an objective and important con to keeping conlangs split across namespaces, which you should already be familiar with: users still won’t know where to look to find the entries they want to see. — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 18:43, 31 August 2026 (UTC)

   "users still won’t know where to look to find the entries they want to see" Better solution to that problem: put all conlangs in mainspace. Whoop whoop pull up ^{♀️ Bitching Betty 🏳️‍⚧️ Averted crashes ⚧️} 07:17, 12 September 2026 (UTC)

   Because of the change to the search feature, the Constructed namespace actually isn't a junkbasket. MedK1 (talk) 23:01, 31 August 2026 (UTC)

   I didn't say it was, I said the Appendix is. Whoop whoop pull up ^{♀️ Bitching Betty 🏳️‍⚧️ Averted crashes ⚧️} 07:18, 12 September 2026 (UTC)

6.  Oppose. jlwoodwa's concerns have brought to light the ramifications of Change 2. Particularly, the external effects, the amount of effort to fix them and many more (minor) things that compile into a bigger issue. They have changed my mind by their convincing mitigation, which is to make dummy entries in the Constructed namespace, acting as soft-redirects for constructed languages residing in the mainspace. From my perspective, this will be a minor inconvenience for the reader, but it is better than creating a huge workload on our editors to make things perfectly consistent. I am assuming that their mitigation would be accepted by the community and even if not, we can revote on the matter at a later time (and potentially with a better solution). TranqyPoo 💬 ✏️ 03:09, 2 September 2026 (UTC)
7.  Oppose I really agree with User:Jlwoodwa's concerns. Moving Esperanto should be a separate discussion. — BABR・kurwa? 19:09, 3 September 2026 (UTC)
8.  Oppose Why keep Eskayan in main ns but not Esperanto? What about Nynorsk and HeBrew? Taylor 49 (talk) 13:48, 4 September 2026 (UTC)

   @Taylor 49 Esperanto is unanimously considered a constructed language by authorities, and Nynorsk/Hebrew are unanimously considered natural languages. Meanwhile, the same authorities’ classification of Eskayan is conflicting: ISO 639-3 and Ethnologue classify it as a natural language, with only Glottolog classifying it as constructed. Hebrew and Nynorsk are unanimously classified as natural. — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 16:33, 4 September 2026 (UTC)

   This is relevant to @Tesco250 as well. See the talk page for more. — Polomo ⟨⁠ ⁠oi!⁠ ⁠⟩ · 16:34, 4 September 2026 (UTC)

   I agree they are classified differently; the bulk of my argument is that the separation by uses is probably more important than by "origins". Tesco250 (talk) 17:54, 4 September 2026 (UTC)

9.  Oppose per my comments above Tesco250 (talk) 14:19, 4 September 2026 (UTC)
10.  Strong oppose MuDavid 栘𩿠 (talk) 02:37, 11 September 2026 (UTC)

Abstain

1.  Abstain, as I am not fully supportive of change 1 to begin with. Juwan 🕊️🌈 09:12, 30 August 2026 (UTC)

---

Proposed votes

The following are proposals for new votes, excluding nominations, in cases where the proposer of the vote prefers that the vote is written collaboratively, or where the vote appears to require substantial revision. If you have not created a passing vote yet, it is recommended that you use this section and actively solicit feedback by linking to your proposal in discussion; your vote may have a better chance of passing if it is first reviewed.

Votes may linger here indefinitely. If changes in policy make a proposal irrelevant, the voting page will be requested for deletion. On the other hand, you do not have to be the creator to initiate one of the votes below. Place any votes with a live start date in the section above at least a few days before that start date arrives.

Forthcoming votes

Wiktionary:Votes/2023-12/Represent the GenAm NURSE and STRUT vowels as schwa

Wiktionary:Votes/2024-06/CFI for mainspace constructed languages

Wiktionary:Votes/2025-04/Deletion of user talk pages

Wiktionary:Votes/2025-08/Prefer Synonyms and Antonyms Templates

Wiktionary:Votes/2025-09/Delete and regenerate pinyin entries

Wiktionary:Votes/2026-08/Artificial intelligence

Wiktionary:Votes/pl-2022-06/Streamlining the approval process of online sources

Wiktionary:Votes/pl-2022-07/Stubifying alternative forms

Wiktionary:Votes/pl-2022-11/Should unidiomatic phrases be included if there is consensus for likely utility to readers?