User:Najawin/Sandbox 10

The Web or: How I Learned to Stop Worrying and Love Deflationism
Ah, The Web. I'm sure we've all thought about it. I know I have. Indeed, I thought about it for quite some time after the R4bp thread, in order to think about whether or not it might actually work for a satisfactory account of validity. I don't think it does, but there are some subtle issues here that we should discuss.

Firstly, is The Web Validity, or is it merely a tool we construct to help us talk about Validity? I think it's clear that The Web of Narrative Connections is clearly a useful tool - if it can be constructed - for any discussion of Validity, since individual authors should have their own views of the DWU in which a subset of their view of the DWU is roughly analogous to a subset of The Web. So it's useful even to someone who thinks that continuity shouldn't touch our validity rules at all, if only for how it impacts the views of others. Scrooge suggests it's the former.
 * I think that by "intended to be set in the DWU", we mean "intended to be part of the Web"

I think this position is... difficult to hold. First and foremost, what about a writer who simply doesn't want anyone to reference their work? A writer that is notoriously disagreeable? They actively discourage people from building off of their storylines whenever possible and try to insist that their work shouldn't loop back around to the rest of what other people are doing, but still write using DWU characters? It sounds almost familiar, if a bit too extreme, no? Is this hypothetical "Angry Harry" still writing for the DWU? Seems to be to me, even if they don't consider their work suitable for others to reference. How about a series that uses some DWU concepts in a completely disjoint way from how they've been used before, (maybe even says that it's "its own universe") perhaps there's some crossovers later on down the line, but there's no backwards narrative connections? (Okay, I'm teasing a bit there, it's not that extreme. But the basic principle applies. What if we take some DWU concepts, rip them free of their original context - you've even suggested that we could do this with an image, a design, to some extent - and then have them run forward in their original series, while they might in the future be fine referencing The Web.) Even if we accept something like The Web, what you're suggesting is too restrictive. But it needs to be this restrictive, because if not, if you allow these dead ends and false starts, then the way you've defined R4 doesn't quite work here.

Alright, now, next, and is it just me, or is this getting really dodgy? Like. Sure. We're dealing with natural language, not formal ones. But when we start trying to define a categorization scheme using the things we're supposed to be categorizing, I get very suspicious. This has historically been a losing play. The insistence of Czech, Tangerine and others to base validity on solely OOU considerations is very well thought out in this regard.

And finally, and perhaps most importantly, I think the issue here is that this is just obviously wrong. The DWU as we mean it on this site is defined by our validity policies. Our article on Doctor Who universe says as much. And these validity policies change over time. It's simply not the case that "Valid" = "part of The Web (not a dead end or false start)". Perhaps, perhaps, this is what it should come to mean, in that we should use The Web as a guide to T:VS. But it doesn't mean that. The two terms are not synonymous, and I don't think it's clear that they even line up at the present time.

Now, nobody is saying The Web isn't important, just that it isn't synonymous with validity. Validity is merely what we make of it. Think of it like the legal/moral distinction. Perhaps we should base our laws off of morality. But legal positivists don't think there's any inherent correlation between them, whereas natural law theorists insist the opposite - "an unjust law is no law at all". I think even if our validity rules didn't correspond to The Web at all, they would still reflect The DWU as this wiki understood it, since we've defined the term to refer to what we construct using T:VS. And I don't think this is a radical position, I think it's the most reasonable one given the plain fact that most people here have lived under a time where certain things we all wanted valid were ostensibly invalid.

Now enough of that, let me explain why The Web isn't important.

Oh, I kid I kid, I jest I jape. It's useful. It's very useful. But I'd be wary to use it so closely as a guide to validity, even aside from the definitional concerns. It's a useful tool, but it should be something we temper with other factors. Why? Well, the same reasons that User:NoNotTheMemes alludes to. And, indeed, this is one of the reasons why I didn't try to offer a Grand Theory of Validity in my opening post. As someone with a copy of Sakurai and Napolitano on my bookshelf not six feet away from me I feel it's my duty to warn Scrooge away from this path with such scant provisions and scant preparation. Avoid its siren song!

At least in addition to this massive web of continuity concerns, we need to think about how our readers will benefit from what we deem valid and invalid, rather than how our editors will. I know some people IRL who are massive Doctor Who fans, just massive. And they despise the wiki and consider it near unusable because it mixes together EU content and show content. I keep telling them that this will never change, that it's a foundational principle of our wiki that all sources are equal, but it really puts them off. And our decision validate all of these R4bp works, well... Some of them frustrated them more than others, but to them it just wasn't helpful, it obscured things more than made it clear. And I don't want to suggest that every reader is like this, but my suspicion is that a lot of them are. One of the earliest conversations on how to deal with validity, which I linked above, Forum:The original inclusion debates, suggested to use italics to demarcate EU vs tv content. And this clearly didn't happen in the long term, though it actually did happen here and there in the early days, people just didn't do it consistently and it fell out of favor. And people have suggested more updated versions since then, see Thread:129501 at User:SOTO/Forum Archive/The Panopticon II. I think there are multiple concerns we need to balance when it comes to validity, these are just the two most obvious. I'm sure, given time, that we can come up with more. Reducing it to one concern is... Misguided in the extreme in my mind. And I certainly wouldn't wish to beg the question against those who feel the need to weigh the scales more heavily towards curating content towards our readers than I do.

Does T:CS work?
Probably not as Scrooge has alluded to it.

Firstly, I want to rebut the idea that R1bp is incoherent or unclear as I've expressed it. I think this is untrue. I think R1bp is relatively coherent and clear, if a bit messy and in need of some tune-up. R3bp is definitely difficult to envision, but this is due to the lack of R3 jurisprudence in particular. Which, well, you contributed to Scrooge, so nyeh. (As I'm writing up the last section, I had a thought. Are Stage Plays generally thought to violate R3 alone? I think there's maybe a chance that one works out there then? But really it seems to me to be usually seen as an R1 issue.)

But can we so easily extricate R4 from the prior work done? Probably not. Certainly we couldn't this time last year - non-narrative fiction was invalid for failing R1 but still covered. We technically still haven't fully relitigated video games and that's an R1 issue. And I think this alone should give us pause. If this 3 step procedure Scrooge is talking about only holds when we've already reformed T:VS to be in line with (what he thinks T:VS would look like after following this same 3 step procedure :P) a version of it that reflects our preferences more heavily than those of past editors, and it simply fails to apply prior to that, well, doesn't that suggest that perhaps this analysis is perhaps too idealistic? It's describing what we wish to happen rather than what actually is happening? If there's no possible way to separate out R1 considerations from past versions of T:VS, then I don't think it's the case than this is what covered and validity mean, and that R4 is a separate stage in the procedure than the other rules. At least, not as of yet! (And, by logical consequence, this means that the various RXbps still come tumbling down like rainfall. If it's all one policy, they're all there.)

And it's worse still, because I'm pretty sure we'll find something floating around that's not a "complete work of fiction" but is covered as invalid even after we fix video games and whatever else we're thinking of. Indeed, Forum:Deleted Scenes and Rule 4 By Proxy seems to suggest this outcome as a possibility.
 * when we consider this, it starts to make a lot more sense that a bunch of deleted scenes lack pages altogether. Stuff which breaks Rule 1 goes un-covered all the time.

I mean, surely this comment in Scrooge's closing post only makes sense if R1 breakers can be covered. (And note that this was after the narrative->fiction shift.)

It's at the very least non-trivial that we can easily divorce R1 from validity in this proposal, and I think probably straightforwardly false that we've meaningfully done so already in a way that R4 truly constitutes a separate stage. (Note, as I'm writing up the last section, I figured out the obvious examples. Stage Plays and Escape Rooms. Arguably they violate R3, but this is still a problem, as they're covered as invalid. I believe it's generally a R1 issue too.) And sometimes failing R1 means not covered, sometimes it means invalid, and it's messy and the entire thing really is one policy, stage three is a substage of stage 1, we assign validity when we create these OOU pages and then create IU pages referencing the labels the equipped sources on wiki use. But I do think there's something wrong with our account of coverage. It's just different from what Scrooge suggests.

Namely. Boy, "not covered" can sure mean a few different things, can't it? And there's not always a clear rhyme or reason as to why something that "isn't covered" is in one category and not the other. Like, sure, validity, invalidity, they pass R2 and R3, and are vaguely speaking works of fiction. To be valid it must pass R1 and R4, to be invalid it has to pass at least part of R1. There are ways in which it can fail R1 and be not covered. These types of stories have pages, plot summaries, and in-universe pages for their characters and plot elements. Sure, the invalid ones don't "talk" to the valid ones, but they're largely the same.

But, like, there are some "not covered" things that we just don't discuss here, for a variety of reasons, failing R2, R3, etc etc. But you can be "not covered" and still have a page here. Or have a page here and even have a plot summary. Or not have your own page, but be a part of a collection page dedicated to a specific type of thing or specific series. I'm really skeptical that we can offload our problems we've found here onto a split between T:VS and T:CS, but dear lord there's a whole hierarchy of how things can "fail to have in universe pages for plot elements" and still be on this wiki that we just have no terminology for and no clear unifying standards. It's very much a case by case basis.

Should we actually split T:VS and T:CS? I don't think so. I think if we want to go down this route, which we may! This seems to be where everyone is headed! If we go down this route, the issues are much deeper engrained, and I don't think that a split this simple will change things. Moreover, I think even this split is so radical that I'd suggest asking every admin we know to be semi-active to weigh in. (And, yes, it's both simple and radical. It's too simple for what it needs to be, and radical for how it changes how the wiki has been operating for a decade.)

If we really want to go down the "reform T:VS" route, my suggestion instead would be that we, again, ask all the semi-active admins to weigh in, just to start. Then the next strategy is that we prepare for quite some time to basically do a full rewrite. We let people know in GB, reddit, twitter, etc, that we're doing a large policy overhaul and want feedback, (I know, I know, I love asking the community. But this in particular is a really big change.) as well as having at least one other person do something like my archaeology project, so that multiple perspectives on the history of the wiki and what's important can be seen as editors read up on this in preparation. I think it's something that we should approach with a lot of care. But this is a suggestion. Once again, people can feel free to disregard it.

Category Theory is a Lie and Alexander Grothendieck Set Mathematics Back 100 Years
So perhaps I'm just too set in my ways, but I don't fully understand Scrooge's response about linear time invalidity for contradiction. His argument doesn't seem even slightly analogous to me. Let me explain why using some handy dandy commutative continuity diagrams. Here's the terminology you need to understand these diagrams, okay? "VN", where N is a number means that this is the Nth "Valid" story we're considering for the diagram, "IN" means this is the Nth "Invalid" story, "V1->V2" means something like "V1 informs the continuity of V2", or "V2 is trying to be in continuity with V1". "V1-/->V2" means the negation of that, and "V1<->V2" and "V1<-/->V2" means that we erase the time dependence of our previous relations.

So back in the days of yore, we had the following two stories, $$ \begin{array}{lcl} & V1, & I1\\ \end{array} $$. You might ask why I'm not including an arrow here. Because, frankly, my argument has no need for it. V1 and I1 might have related continuities, they might not. (EG: It's hard to deny that Shalka is continuous from the classic series.) My argument does not assume that invalid stories and valid stories are intentionally discontinuous. Indeed, it considers the entire issue a red herring. Now, given these two stories, we can consider another story that references the two of them. $$ \begin{array}{lcl} & V1 & \rightarrow I2 & \leftarrow & I1\\ \end{array} $$. Now, I2 was made invalid, and this, crucially, was because of the connected arrows, because of a pattern present in this diagram, namely, $$ \begin{array}{lcl} & I1 & \rightarrow & ? & \Rightarrow & I1 & \rightarrow & I2\\ \end{array} $$. The sequels/prequels to invalid stories thread changed this. Now, depending on factors other than patterns in these diagrams, both $$ \begin{array}{lcl} & V1 & \rightarrow I2 & \leftarrow & I1\\ \end{array} $$ and $$ \begin{array}{lcl} & V1 & \rightarrow V2 & \leftarrow & I1\\ \end{array} $$ can exist. The R4bp thread concluded with the idea that considering diagrams of the form, $$ \begin{array}{lcl} & V1 & \rightarrow V2 & \leftarrow & I1\\ \end{array} $$ we can replace them with $$ \begin{array}{lcl} & V1 & \rightarrow V2 & \leftarrow & V3\\ \end{array} $$. This is asymmetric. Scrooge suggests the issue is one of looking at the same graph while removing the time element. I wish to submit that this is clearly false. If we're considering diagrams of the form $$ \begin{array}{lcl} & V1 & \rightarrow \quad ? & \leftarrow & I1\\ \end{array} $$ and simply deciding whether "?" is to be valid or invalid, I don't see how changing this diagram to $$ \begin{array}{lcl} & V1 & \leftrightarrow \quad ? & \leftrightarrow & I1\\ \end{array} $$ changes the calculation one whit. (Indeed, it's arguably because of this change that my argument works! We need there to be a symmetry between going from $$ \begin{array}{lcl} & V1 & \leftarrow & I1 & \Rightarrow & V1 & \leftarrow & V2\\ \end{array} $$ to $$ \begin{array}{lcl} & ? & \leftarrow & I1 & \Rightarrow & I2 & \leftarrow & I1\\ \end{array} $$ and there just isn't without assuming time invariance. My argument really was that using these arrows at all was circular reasoning, given how our diagrams were populated, but time invariance actually makes it stronger. Otherwise this second form of the argument only works on diagrams of the form $$ \begin{array}{lcl} & V1 & \rightarrow I2 & \leftarrow & I1\\ \end{array} $$ which is a rather different issue.)

Perhaps the difference will become more apparent if we chart out his proposed counter example and why I don't think they're similar?

Scrooge suggests we consider the following diagram instead: $$ \begin{array}{lcl} & V2 & \nrightarrow & V3 \\ & \uparrow & \nearrow \\ & V1 \\ \end{array} $$, or, if you prefer, $$ \begin{array}{lcl} & V2 & \nrightarrow & ? \\ & \uparrow & \nearrow \\ & V1 \\ \end{array} $$. And this is clearly non analogous, right? I mean, even if you insist that my hypothetical has to populate the continuity relationship between V1 and I1, a view I straightforwardly reject, even if you apply time invariance to this graph as well (bear with me, this is actually the one issue here, diagonal arrows can't do this on mediawiki, so I'm going to turn it into a box diagram) $$ \begin{array}{lcl} & V2 & \nleftrightarrow & ? \\ & \updownarrow && \updownarrow \\ & V1 & = & V1 \\ \end{array} $$, in this analogy, there's still a giant hole - that both stories already propagating the diagram are already valid, and the "non continuous" arrow is a relationship between one of them and the new story we're considering, not between the previous stories that already are present in the diagram.

Now, one might argue, perhaps we shouldn't privilege past states of the diagram over future ones. That if we fill in those continuity arrows between I1 and V1 from my example we get $$ \begin{array}{lcl} & V1 & \nrightarrow & I1 \\ & \downarrow & \swarrow \\ & ? \\ \end{array} $$ and then adding time invariance to this diagram gives us $$ \begin{array}{lcl} & V1 & \nleftrightarrow & I1 \\ & \updownarrow && \updownarrow \\ & ? & = & ? \\ \end{array} $$ and then when we realize that in his example we clearly agree with validity for the question mark, so we do so here as well, and then this back propagates to I1. So $$ \begin{array}{lcl} & V1 & \nleftrightarrow & V3 \\ & \updownarrow && \updownarrow \\ & V2 & = & V2 \\ \end{array} $$. Potentially then all structures of the form $$ \begin{array}{lcl} & X & \nleftrightarrow & X \\ & \updownarrow && \updownarrow \\ & X & = & X \\ \end{array} $$ are either all valid or invalid?

Perhaps this is what Scrooge is suggesting all along? It's not time invariance wrt media, it's time invariance wrt the decisions we make about these pieces of media. I don't really think he is? I'm just trying to steelman this argument. I don't really see a reading in which the two are analogous otherwise. Like - it's got some weird implications for how we conduct validity debates at the very least. At this point, not being a category theorist, and that's probably the best way to think about this, I think I have to beg off discussion ever so slightly, but my intuition is that this is precisely the reductive version of R4 Scrooge has suggested and people have not largely been fans of. It's that instead of springboarding off of Unearthly Child our web springboards off of every piece of DWU media. That's my guess. (I think this is the first time I've ever wished Amorkuz was here - I think he was an algebraist?)

Regardless, perhaps Scrooge can clarify what he means? I really don't see any similarities - I don't see how this is in principle a rebuttal.

Continuity, References, Android Boyfriends, oh my!
And so we come to this. Perhaps the heart of Scrooge and my disagreement, and perhaps the thing we'll simply never see eye to eye on. I think I can convince him on much of the rest that's gone before, or perhaps that we'll come close to mutual understanding. But this, no, I think we're just speaking fundamentally different languages here. Case in point
 * The question of whether the Doctor's line in Time of the Doctor is referencing the Shalka Master or not… already exists. It would exist, just as irresolvable, even if we decided to throw up our hands and delete the #Continuity sections. Heck, let's also delete #References while we're at it. The question is still there, laughing at us, because — I cannot believe I did not spell this out sooner — we still have to decide whether we put that line on The Master (Scream of the Shalka) or create a separate the Doctor's boyfriend (The Time of the Doctor) page.

We do the second. It's site policy. We do the second. This has been upheld time and time again. We do not go further than what the source tells us. Indeed, I state something very similar to the inverse of this argument in my OP.
 * We all agree that for various events in the DWU competing events are relatively common, yes? And it's important to report neutrally on these accounts, stating what each source tells us, and not to speculate further. So if in one source we see precisely X and in another Y, we say that in one account X was held to have happened and in another Y was held to have happened, refusing to speculate further, refusing to say further than what X tells us, and refusing to say further than what Y tells us. And previously there was some symmetry between how we handled these cases and invalid sources and valid sources referencing invalid ones. An invalid source says X, but it also, implicitly, says that it cannot be trusted and we can't use it to write articles, so we ignore it, and then a valid source says precisely Y, so we say Y. But now we've decided to break this symmetry.

And this is where I feel our intuitions just differ fundamentally. Mine, vs yours and Nate's and probably many other's. For me it's less than ideal to have these pages for the things we all clearly know are The Master, etc, on pages other than theirs. But I'd rather make sure that we don't speculate over having a flawed rule in order to merge them. (And lest someone suggest that any rule we create to merge them will be fine because "we'll only apply it to pages we know concretely should be merged", first of all, I have a bridge in Brooklyn to sell you, second of all, you can imagine how little this is going to convince the mathematician. "Oh, we have this cool machinery that's proving all these theorems we're pretty sure are true. P!=NP, Collatz, Navier Stokes, Golbach, Twin Primes..." "Okay, but I've seen the machinery, it's pretty suspicious and it looks like it can be used to imply basically anything." "Yeah, but that's okay, we're only using it on stuff we already know is true.")

Now, I'd like to rebut the idea that I'm against "continuity" or "referencing" at all. This isn't true. I just don't think that we, the wiki editors, have access to that knowledge. (Because, again, actual intentionalist. Well, partially. I just think that textual evidence is a very poor guide for what precisely we're looking for here and it's going to lead to incredibly messy differences of opinion. We are epistemically limited in this matter to a truly remarkable degree.) And so if we decide to use it for validity it must be based on authorial statements. As I said in the original R4bp thread. I do think it's occasionally possible for there to be relatively unambiguous and universally understood statements of the type of authorial intent we're looking for in a text. I just believe they're incredibly rare. And this is what the Android Boyfriend example shows to me - not the impossibility of communication between individuals, but that we fundamentally cannot be trusted as to what "clear" actually means when we're reading a text. One person sees a particular reading as obvious, another sees another reading as obvious. It might still be the case that in a different section of the book everyone, or near everyone, reads the text in the same way. But we should be very skeptical that our readings of texts are the sorts of things that we need for this policy to work without authorial statements. We're just so epistemically limited here. (I also note that I don't wish to define the DWU as R4 intends it as completely divorced from continuity. It is defined that way, well, modulo R4bp. It's constructed by T:VS.)

But no, I don't think this line of attack is particularly compelling. It's explicitly site policy to not speculate, to only say what the source tells you, nothing more. I'm quite comfortable declining to do so, or at least trying my best not to, (because I know I've done so in the past, forgive me Scrooge, for I have sinned) and being deeply skeptical of those things that I accuse of subjectivity. I note as well that this criticism rests on trying to dissolve the distinction between the language and the metalanguage, as it were, how we add things to and categorize them within the DWU, and how the DWU resolves itself once things have been added. I am, of course, deeply skeptical of this.

And so we end on the sky not falling
I think I'm rather more cynical than Scrooge on the subject of whether people will revisit prior subjects that have been validated using R4bp, but I look forward to being proved incorrect. In a way this section of his neatly reflects on a comment OS25 made a week ago.
 * The big reason I feel so strongly about Rule 4 By Proxy is not only that it makes sense, but also that it’s essentially only made our site better in every case that it’s been used.

Well, I don't think it makes sense, and I think this confidence is misplaced, given the closing of one recent validity debate (there but for the grace of Rassilon...) and the opening of another. R4bp can make the site change in ways that I think many of us are not necessarily fans of. So I'm not certain that we should be appealing to consequences here. I think they're very much open ended. For me, as I said above, the reasoning comes first. I can live with Ninth Doctor 4 (The Tomorrow Windows). It's irritating, but not ideal. I can live with a prominent, often referenced invalid story. I can live with weird implications of well thought out rules consistently applied. Poorly thought out rules, or well thought out rules inconsistently applied? That's when I have real frustrations. And R4bp is the last category of the three. (Well, it's well thought out in its own right. I don't think how it jives with the rest of T:VS was quite as thorough.) By its very nature, as it's currently written it's arbitrary in application. Arbitrary in terms of whether we decide to use it or something else? I can live with that - that has to happen to some extent. But in how we apply it? Not so much. And what's more, we don't have to. We're biting a massive bullet here for no reason at all that I can see. Quite a few.

I've sat here trying to write about my frustrations of us, all of us, perhaps just not communicating and not quite having the same ranking of where things fit in our priorities for about 20 minutes now. And I'm just not seeing it. Everything I write borders on hysterical, far too verbose, or cruel. Mainly hysterical. Maybe we'll just never bridge that gap. But I hope we can. I truly hope that we can come to an understanding on why things being arbitrary in terms of application in particular is such an issue.