PathfinderWiki talk:Canon policy/Archive 4

Fiction tier

Changes Accepted This section contains a discussion about changes to this policy that have been accepted.

I have just noticed that, although the fiction lines are mentioned in the text of this policy, we have no official tiering of Pathfinder Tales novels or the web fiction. I think they should be added as they do enrich our wiki considerably.

At Paizo's fiction page they split fiction into:

• Tales: Short Fiction, Novels, Web Fiction
• Pathfinder Legends
• Pathfinder Comics

In reverse order, Pathfinder Comics are already housed in Tier 4 and I suggest no changes.

Pathfinder Legends: 'All licensed Pathfinder campaign setting material in non-Paizo sources' nicely covers Pathfinder Legends and places them in Tier 4. For clarity, we could call them out there as a specific case though. This is the first change suggested.

Tales: although they are grouped together at Paizo, we should consider them separately even though, in my initial opinion below, they can be handled as equivalents. The 'Short Fiction' and 'Web Fiction' are practically the same thing and both were published by Paizo. The 'Novels' were published by Paizo and by Tor but, content-wise, there is no difference, I'd say, in canon level based on that switch. I also feel that there is no difference in canon level between the 'Web Fiction' and 'Novels': the stories often concern the same fictional characters. So, I'd argue, 'Tales' may be handled as a unit.

For tiering, I would say that 'Tales' are a minimum Tier 3 source, but will argue here for Tier 2 inclusion as much of what we know about various areas of Golarion and some of the famous faces comes from this fiction. This is the second change suggested. -Fleanetha (talk) 21:38, 4 August 2018 (UTC)

Agreed. As I understand it, even when TOR was printing/distributing, the tales line was being edited in-house at Paizo. Along the same lines the same editors worked on the web fiction as the book fiction, so I think they should be clumped together. I also think they are just as valid sources as Campaign setting books. So basically, I'm just agreeing with everything you said. Edit: after reading the discussions about Ostovite, the unreliable narrator can cause issues... Maybe we should add a 2.5 tier due to the narrator? "sources with an in world narrator such as fiction are dropped 1/2 a tier to prevent confusion if this narrator is misinformed"-- Cpt kirstov (talk) 14:09, 6 August 2018 (UTC)

I agree with Cpt kirstov that the limited fictional character perspective of Tales novels, and fiction in general, makes them sometimes inherently less reliable as canon sources than the omniscient voice of Tier 2 Campaign Setting, Modules, Player Companion, and PFS scenario works. I also think we should specifically include Pathfinder's Journal fiction in APs in this decision, and at the same level as Tales fiction, for the same reasons.

• I agree with Legends being a Tier 4 source.
• I agree that Tales should be at least Tier 3. I would argue for them being in Tier 3, but I also do not oppose Tier 2, on the condition that when resolving otherwise irreconcilable conflicts between works of fiction and sourcebooks, the voice and authority of the in-universe source should be strongly considered, and very rarely in its favor.
• I propose specifying Pathfinder's Journals at the same tier as Tales. -Oznogon (talk) 21:48, 8 November 2018 (UTC)

I like the way this is going. I don't think we are at a resolution stage yet but, I think, introducing a Tier 2 Fiction level might be a useful way forward. That would pick up the 2.5 level suggested by Cpt kirstov for valid reasons, but also incorporates all fiction as proposed by Oznogon. It raises fiction above Tier 3, which feels right to me, but allows a more formal Tier 2 source to take precedence on the occasion of a conflict—in fiction, an author may want a protagonist not to be knowledgeable about a certain concept, and we need to account for that. Having 'Fiction' in its name means it should be pretty clear to what we are referring at that level too. I formally amend my proposal change for this to be the way forward now. --Fleanetha (talk) 23:15, 8 November 2018 (UTC)

This is actually coming up On the messageboards right now. I came here to link to how we categorize them, and found this policy change was never settled. I officially propose Legends be tier 4, and Tales and AP fiction as tier 2, as the unreliable narrator is no more an issue than some of the retired PFS scenarios, and they are still tier 2. -- Cpt kirstov (talk) 17:03, 15 November 2018 (UTC)

Proposed Text changes:

(Change in bold)

Tier 1

• All Pathfinder Adventure Path volumes excluding fiction in Journal section.

(Addition)

Tier 2

• Pathfinder Fiction produced or edited by Paizo Inc.
  • Includes Pathfinder Tales, Web fiction, and Journal section of the Adventure Path volumes.
  • If a Fiction source conflicts with another tier 2 source, take into account the possibility of an unreliable narrator in fiction.

-

I approve! Thanks, User:Cpt kirstov! -Oznogon (talk) 17:20, 15 November 2018 (UTC)

Just before we approve this, I have found earlier discussions about Fiction and canon tiering in the archives to this page, here: PathfinderWiki talk:Canon policy/Archive 1, and here: PathfinderWiki talk:Canon policy/Archive 2. We should probably be aware of what was said by the community about this subject before. Fortunately, it seems supportive of the conclusions being drawn here. That latter archive should probably get a 'concluded' / 'See also' tag attached when we complete this exercise, as it specifically deals with Fiction. Thus, we have taken just over eight years to conclude this; I don't think that is actually a record. Thanks for the Paizo message board link too Cpt kirstov. --Fleanetha (talk) 17:57, 15 November 2018 (UTC)

Summarising where we are:

first change suggested: no one has disagreed and all contributors here have agreed that Pathfinder Legends are rightly placed in Tier 4, though they should be called out as such formally in the policy.

second change suggested: All fiction (Short Fiction [which includes AP fiction and web fiction, but happy to call AP fiction out specifically for clarity], Novels, Web Fiction) is looking like a Tier 2 candidate now. All contributors here have considered a Tier 2.5 construct but it looks like, for simplicity, full Tier 2 is now more favoured. I am happy with that, as were the older contributors in the archive, so I am [stay with me] reverting my amendment above back to the original second change proposal of all fiction being called out as being Tier 2.

User:Cpt kirstov, Oznogon & Fleanetha have agreed this, I believe, reading above. I suggest we time box this now and, if no dissent is raised, this will become a formally accepted policy change seven days hence at 7pm GMT, Thursday 22 November 2018. No one should feel that this is now set in stone: if you disagree with this proposal change, please join this conversation in the next week please. All opinions are valid and we'll just reset the clock. --Fleanetha (talk) 18:25, 15 November 2018 (UTC)

Changes accepted

We shall implement these changes now to PathfinderWiki:Canon policy. Thanks to all who have helped shape this change. --Fleanetha (talk) 20:17, 22 November 2018 (UTC)

Policy enforcement (categories)

Changes Accepted This section contains a discussion about changes to this policy that have been accepted.

“

In the case of an irreconcilable conflict between sources, the source from the most authoritative tier should be given preference (in the case of an intra-tier conflict, the most recently published source should be used, excepting extenuating circumstances or consensus to the contrary).

”

So this is apparently how canon conflicts are resolved on this wiki, but I've noticed that this particular bit seems to be only enforced within an article's content itself, not by categories (which seem to take all sources as canon with no preference to any of them). Take the Baphomet article: the infobox only displays CR 27 (with a footnote explaining how it superseded the old CR value in PF18), but the article itself is still placed in both the CR 27 and CR 31 creature categories, even though this individual CR value has been superseded, and CR 30 is the cap in PFRPG after the game moved on from 3.5E (as stated by a few Paizo employees numerous times). And even then, the enforcement of this unwritten rule looks rather inconsistent (like how Achaekek was never re-added to Category:Hell/Inhabitants for example). For the record, since the categories seem to take all sources as equal, should the Red Bishop article be added to Category:Chaotic neutral inhabitants as well (even if the CN alignment in the source material has been confirmed as a typo)?

I think the wording of this policy should be amended to better reflect how it is actually enforced in practice. - HTD (talk) 15:45, 8 November 2018 (UTC)

Could you suggest the form of wording you think would be a better substitute, HTD, and why? --Fleanetha (talk) 16:20, 8 November 2018 (UTC)

Maybe keep the information in the quote and add information about categories? so it would look like:

“

In the case of an irreconcilable conflict between sources, the source from the most authoritative tier should be given preference (in the case of an intra-tier conflict, the most recently published source should be used, excepting extenuating circumstances or consensus to the contrary). When this is done ensure the categories are updated properly.

”

In theory, we could also do a subcategory for each category that is removed due to conflicts. For the Baphomet examples above it would be something like Category:CR 31 creatures with conflicts if we wanted to keep the categories on the page without damaging cannon. -- Cpt kirstov (talk) 17:42, 8 November 2018 (UTC)

Categories are metadata provided by the Mediawiki software and are designed and primarily used for discovery. They are not intended to reflect the current state of canon; the article content is. We can state that explicitly in the canon policy to eliminate any confusion.

Beyond this, there are a few other points I want to stress:

• Game mechanics are not canon content for the purposes of this wiki. This is described in more depth in Project:No crunch, and inferred in Project:Scope of the project and PathfinderWiki:Point of view. We offer categories to help people discover articles, but for categories based on mechanics, the canon policy is not as relevant. Again, stressing in this policy that categories are not intended to reflect only the current state of canon would help. A standalone policy specific to the purpose and usage of categories on this wiki could also help.
• Assume no conflict. Mechanics and the canon state of the setting are updated over time, but those updates don't invalidate the past unless the conflicts are irreconcilable. This includes explicit confirmation of the current mechanical state of a subject. Again, because categories are intended to aid in discovery, there is value in continuing to categorize content based on how it was described in the past. The article content should still reflect the current canon state, but documenting irreconcilable conflicts both helps users understand the change and more completely and accurately documents the history of the subject. Making this explicit in the canon policy could help avoid further conflicts between editors.
• Categories are not inherently exclusive of each other. A character described as level 20, for example, can also be described as being of level 20 or greater, especially if canon sources have previously described the character in both ways. A character who is undead is deceased even if their essential form has not been destroyed—we have separate {{Person}} infobox fields to help document the distinction, for example. Overcategorization and overlapping categories are not inherently harmful. If an article's categories are accurate but confusing, that's an opportunity to more clearly describe the situation in the content, not a problem with the categorization.
• Newer sources should supplement, not replace, older sources as citations. Concurrently with how we assume no conflict as a baseline, we should assume that both older and newer sources are valid for canon content. It's clear that 3.5 ruleset mechanics are being even further deprecated in recent PFRPG sources, but unless we have explicitly resolved conflicts, we should not remove 3.5 sources as citations for non-mechanical canon content. Even when there is an irreconcilable conflict, we should document it and cite the source on the Talk:Subject/Conflicts page.

I'm not in favor of describing conflicts in category names:

• For categorization, we already have Category:Canon conflicts.
• For editors, we can already include comments in the wiki source (<!-- This category is included because ... -->)
• For everyone, we can already cite conflicts in the article to describe the situation ({{Conflict}}, Talk:Subject/Conflicts). Finding an appropriate way or place to implement {{Conflict}} to best reflect this is a per-article concern best hashed out on Talk pages, not in a policy discussion.

Optionally making the presence and description of conflicts more visible, similar to how we use {{Badges|spoiled}} and the spoiler warning gadget for configurable visibility of spoiler warnings at the top of articles, could have potential benefits not limited to categorization for people especially concerned with conflicts, while allowing people without those concerns to more easily ignore them. This can happen without a canon policy change and without deleting any categories, citations, or content. -Oznogon (talk) 20:59, 8 November 2018 (UTC)

I propose adding this section to Project:Canon policy as a second-level heading:

Categorization

Categories are metadata provided by the Mediawiki software and are designed and primarily used to help chroniclers and visitors discover related content on the wiki. They are not intended to reflect the current state of canon; the article content is. This can include adding categories reflecting the past state of a subject, such as older game mechanics or changes over time to that subject that result in a change in its categories. Note that production errors, such as Old-Mage Jatembe's alignment misprint in Inner Sea Magic or the "buttery knife" in Adventurer's Armory, are exempt from this; superseded content can be categorized, but content that was never deemed correct should not be.

For example:

• A fifth-level fighter who dies over the course of the canon's timeline does not lose its Category:Fighters and Category:Fighters of 5th level categories when it gains the Category:Deceased inhabitants category, even though that character ceases to exist as a fighter, because that character still exists in some canon works as a fifth-level fighter, and a visitor looking for fifth-level fighters should still be able to find an article on that character in Category:Fighters or Category:Fighters of 5th level. The fighter's {{Person}} infobox can still describe the character by class levels, but should include a deceased or destroyed date, year, or event as appropriate. Because class levels are crunch, they should not be included in the article's content.
• A creature depicted as being CR 20 in an older source does not lose Category:CR 20 creatures when a newer source updates it to CR 24 and results in the addition of Category:CR 24 creatures, because that creature still exists as a CR 20 creature in some canon works and should be discoverable in either category. The infobox should state the current canonical state of the creature, so its infobox should display CR 24, not CR 20, and {{Conflict}} should be added to the CR value in the infobox to describe the change. Since the CR is crunch, it should not be included in the article's content.
• A city that was a metropolis in a source published in 2016/4716 AR and was destroyed in a source published in 2017/4717 AR might gain Category:Ruins but should not lose Category:Metropolises, because that city still exists as a metropolis in some canon works and should still be discoverable in Category:Metropolises. Its infobox should depict the city's current inhabitants, if any, and its history should explain that the city once had a population large enough to be classified mechanically as a metropolis, citing the older source.

Also, I propose implementing a gadget similar to the spoiler warning to allow editors to opt in to more prominent messaging about conflicts on a page. -Oznogon (talk) 20:59, 8 November 2018 (UTC)

HTD, concerns about the categorization on Achaekek or Red Bishop should be raised on Talk:Achaekek and Talk:Red Bishop, respectively. It would help to cite any additional sources on those articles' talk pages that suggest confusing or unexplained changes or categorization. -Oznogon (talk) 20:59, 8 November 2018 (UTC)

Nicely put Oznogon and hear! hear! I think I'll add one not insignificant point here too. Regardless of canon, our editors have spent time and effort crafting material for the wiki in order to help others based on canon material. This is important and valuable work. We should remember this before we ever delete any material; this is in addition to all of Oznogon's points above.

We are likely to see many changes ahead with a new edition of the game, but what a chronicler of Golarion would write today in 4718 AR is still valid information ten years hence, even if it is no longer an accurate reflection of the then current status in 4728 AR. We tweak, not delete. --Fleanetha (talk) 23:37, 8 November 2018 (UTC)

I renamed this subsection to clarify that the discussion is mostly about enforcement regarding categories.

My opinion is that we should be as clear as possible with categories. If that means that we include more granular categories, I'm good with that. That said, with the forthcoming edition change, we are likely to encounter a TON of slight changes in mechanical categories that won't be reflected by in-world canon. For example, if the changes to the paladin class from the most recent Pathfinder Playtest update make it into the final game, we could see existing characters in the world re-done as non-LG paladins in 2E. Or Seelah could become the iconic knight, if paladin is a subclass of the knight class (or whatever it gets called). Some spells and monsters are going to change CR or level (maybe?) and so on. None of these are things that someone in-world would know, since spell levels, class levels, and monster CRs are all crunch abstractions that don't exist in-world. We're already, then, going to have twice as many categories for these subjects reflecting both editions of the game. Add to that even more categories because we're reflecting aspects of the subjects that have been deemed not canon, whether due to retcons or initial errors, and we risk having more categories than people can easily navigate.

As for the proposed addition to the policy, I don't have issue with any of that text, and think it would be a fine addendum to clarify categorization best practices. I'm not sure it needs to be a policy, however, and think it might be a better fit as a guideline (the lack of differentiation between the two on our wiki notwithstanding). Until we separate the two, which is the topic of a different discussion), I'm fine with adding it as written.— Yoda8myhead (talk) 00:46, 9 November 2018 (UTC)

I think I'm seeing a minor (or not-so-minor) difference between the second example and the other two. The first and third are examples of story progression, as in (from an in-world POV) this city is now destroyed, but when it existed it was a metropolis; or this character is now dead, but when alive he was a level 5 fighter; etc (the new info is now correct, but the old info was also valid some point in the past of the campaign setting)... In contrast, CR changes (as well as some other changes, like Ileosa's bard levels) are not accompanied by any advancement of the story, and the whole reason is that PF1E is balanced differently from D&D3.5E (except in cases like Khorramzadeh, where there's an in-story justification for the change). In that sense, this example has more in common with the change in Achaekek's divine realm; as in, for all intents and purposes, the new info has completely superseded the old info (for the record, the Achaekek and Red Bishop examples are only here for the purpose of illustration).

Speaking as a person who consults the wiki, I was rather confused by how the CR 30 category listed both individuals whose current CR is 30 as well as those who were listed as such in PF18 (my thought was something like this: Paizo employees have stated that these CRs are no longer valid; why are these categories still here?). From what Yoda said about the category name clarification, I guess we could call that category something like 'Category:CR X creatures (superseded)' for creatures whose CRs have been updated.

As for 2E, seeing how most creature CRs (or rather levels) have been kept unchanged in the playtest bestiary, we probably won't need twice as much CR categories as we have now; I think we can just keep the current categories as they are currently (replacing CR with whichever term PF2E uses) and keeping the 'CR X creatures (superseded)' categories to accommodate the rare few whose CR changed. - HTD (talk) 01:14, 9 November 2018 (UTC)

I fully understand the points you're making about the validity of creature CRs in the Pathfinder RPG rules. I understand the confusion you have about it. I understand the points you're making. You opened this proposal asking for changes to the wiki's policies to clarify their intent and align them with administrative practice. We are trying to do that. Instead of arguing that the proposed intent is unclear, you're arguing that it is incorrect, which is why this disagreement has been so persistent and contentious across article and user talk pages and forum threads.

CR changes (as well as some other changes, like Ileosa's bard levels) are not accompanied by any advancement of the story, and the whole reason is that PF1E is balanced differently from D&D3.5E

I am not sure how to more clearly or kindly communicate to you that this point is not relevant to categorization on this wiki. If this is a completely intractable point for you, this might not be the best wiki for you to contribute to.

I've explained why in my response: game mechanics are not content on this wiki, categories are not content, categories aid in discovery, categories can and should reflect changes to aid in discovery because they are not content, and categories should not be inherently exclusive of each other. These should not be new points—Fleanetha and I have made them to you, repeatedly and directly in plain terms via multiple channels across article and user talk pages and forum threads.

Paizo employees have stated that these CRs are no longer valid; why are these categories still here?

As you yourself stated for the other two examples that don't mention CR, but cover the same ground:

the new info is now correct, but the old info was also valid some point in the past of the campaign setting

For the purposes of categorization, this is exactly the point I and Fleanetha have been making on this subject. The CRs were valid at some point in the campaign setting. For categorization, I do not care if they are invalid now. For categorization, I do not care if they are fundamentally made invalid by the PFRPG rules system. For categorization, if something was described as one CR in a 3.5 book and a different CR in a Pathfinder RPG book, I don't care why. It goes in both categories so someone who opens a book from either era and sees a creature is of a CR can find other creatures on the wiki also described as having that CR. Removing, or moving that category to one that only includes "superseded" decisions, removes that functionality.

And per your illustration that Cat:Hell/Inh was not "re-added" to Achaekek, it was never added in 10 years of revisions to that article. It was an unintentional omission that you've been wielding in this debate as though it's meaningful or intentional, and Inner Sea Faiths made it irrelevant anyway by stating that he has a lair in Avernus—making him for the purposes of categorization a current, canon inhabitant of Avernus, even with the clarification by Jacobs and in Planar Adventures.

Most of these decisions require a similar level of subjective, case-by-case, collaborative effort—a policy on this wiki will not and cannot cover every edge case. The default should be to ask before making a unilateral change to delete categories someone else added, which has been the biggest related point of contention beyond this specific debate.

From what Yoda said about the category name clarification, I guess we could call that category something like 'Category:CR X creatures (superseded)'

Yoda said, emphasis mine:

Add to that even more categories because we're reflecting aspects of the subjects that have been deemed not canon, whether due to retcons or initial errors, and we risk having more categories than people can easily navigate.

I do not see any way how creating completely new categories for edge cases instead of using existing categories—in either case affecting no more than a fraction of a percent of the articles on this site—will help avoid this concern of Yoda's.

The discussion about 2E is a different discussion and more broadly relevant, because it has definite, already stated canon non-mechanical ramifications to the timeline. It's also a separate discussion, and too early to have because it remains in flux. -Oznogon (talk) 02:36, 9 November 2018 (UTC)

(reset indent)

All right, so it looks like I misunderstood the real purpose of the categories on this wiki then. I guess we could add that clarification about categorisation to the canon policy now. - HTD (talk) 04:11, 9 November 2018 (UTC)

I've revised the proposal to exempt production errors. The added text is:

Note that production errors, such as Old-Mage Jatembe's alignment misprint in Inner Sea Magic or the "buttery knife" in Adventurer's Armory, are exempt from this; superseded content can be categorized, but content that was never deemed correct should not be.

If there's no additional comment on that addition or this proposal in 7 days (16 Nov 2018), I ask that we officially approve this policy revision. -Oznogon (talk) 02:20, 10 November 2018 (UTC)

Marking as accepted and implementing these changes. -Oznogon (talk) 01:10, 20 November 2018 (UTC)

Canon tier of playtest materials

Changes Accepted This section contains a discussion about changes to this policy that have been accepted.

I've noticed that some creatures from the Pathfinder Playtest Bestiary are being documented (today: sewer ooze). This is fine, but I propose that:

• Canon content in playtest materials be added to Tier 2, per James Jacobs and Mark Moreland, to reflect their intent.
• Mechanical content in playtest materials be added to the list of non-canonical sources, reinforcing Project:No crunch and clarifying that playtest rules are intentionally in flux.

This doesn't stop and shouldn't discourage new content from 2E playtest material from being documented, but reinforces that playtest materials can and should be overruled by current Tier 1 and future published Tier 2 canon works. -Oznogon (talk) 01:25, 20 November 2018 (UTC)

Agreed - sounds sensible. --Fleanetha (talk) 12:24, 20 November 2018 (UTC) Sorry User:Fleanetha, please feel free to remove this comment and strikethrough or amend if your agreement still stands w/r/t today's revisions to the proposal. -Oznogon (talk) 21:36, 27 November 2018 (UTC) Agreed - don't worry. Sounds sensible rewording post the clarification we've had. --Fleanetha (talk) 22:17, 27 November 2018 (UTC)

For an active use case, see The Last Theorem, updated with content from playtest adventure Doomsday Dawn by HTD that expands and advances its role in the setting and potentially conflicts with minor mechanical elements of its description in The Pact Stone Pyramid. -Oznogon (talk) 18:46, 27 November 2018 (UTC)

From a conversation with James Jacobs about this topic, specifically raised because of the Last Theorum issue mentioned above, he said that Doomsday Dawn should be treated the same as any other Pathfinder Module. Thus, it should supersede contradicting information from older modules, like The Pact Stone Pyramid. He said that part of the intention of changes to canon elements in Doomsday Dawn was to course correct strange elements from early adventures when the Modules and AP teams were pretty siloed and not working toward a cohesive setting. Since this was a conversation in the Paizo offices, this post can be considered a Tier 0 clarification of intent from Paizo staff (as made public by me).

Addendum: Thus, I propose a change to the proposal: Canon content in playtest materials be added to Tier 2. Mechanical content, I would argue, should fall under the No crunch policy and not be considered canon at all (especially because all mechanics in the playtest are so transitive and never intended to be the final form of the rules). Yoda8myhead (talk) 21:22, 27 November 2018 (UTC)

Agreed, and I've updated my proposal to reflect those suggestions. Thanks for following up on this, Yoda8myhead! -Oznogon (talk) 21:33, 27 November 2018 (UTC)

Agree - Cpt kirstov (talk) 23:46, 27 November 2018 (UTC)

I'll put a stake in the ground here: if there's no further discussion or objections in 7 days (December 6), I'll resolve this proposal as approved and implement it in the policy. -Oznogon (talk) 00:36, 30 November 2018 (UTC)

Proposal approved

Implemented in the policy and marked as approved. -Oznogon (talk) 19:35, 8 December 2018 (UTC)

Canon state of setting changes occurring during an AP

Copying over this comment from User:Brandingopportunity on PathfinderWiki talk:Nocticula Revision Initiative:

We should probably include a discussion on defining the wiki policy for major canon updates like Nocticula: they only happen once we have "proof" that they are indeed permanent in subsequent publications. This mostly has to do with APs updating the "base canon" of the setting. We should also note in this discussion that the "base canon" will be wholly updated once Pathfinder 2: Electric Boogaloo is released. --Brandingopportunity (talk) 01:22, 14 November 2018 (UTC)

I have no proposal for this, but the discussion never materially happened on NRI, in part because Fleanetha is avoiding that project due to spoilers. -Oznogon (talk) 18:13, 20 January 2019 (UTC)