On Umineko

15 Jul, 2026

Who are you? I think there are two types of people here. I will write to each of you separately. The first audience is Umineko-curious. You've seen the memes of it and have seen it near the word "peak" enough times to enter your Someday Queue and you're reading this to see if you should adjust it up or down. I will write to you first, about the story from the outside. The second audience has completed it and wants to read thoughts about the story from the inside.

To The Curious

Umineko is about everything. It is most literally a murder mystery; a rich family convenes over an inheritance struggle on an otherwise uninhabited island during a storm, they encounter some eerie signs of the supernatural, and they get murdered one after another, while the remaining people gather clues to find the culprit. Thematically, though, it's about the subjective nature of truth. While two people might look at the same scene or the same set of facts, they can draw wildly different conclusions based on their life experiences and their feelings about the people involved.

The subjective slippage around truth often happens around situations big and small in the present, but it can happen across time, as well. While reading Umineko, I've also been reading Reconstruction by Eric Foner. In its preface, he discusses how views of Reconstruction evolved. The first historical view of it was the Dunning School, early in the century. In this telling, the defeated South accepted defeat and were eager to integrate back into the country, but the ruthless Radical Republicans and the childlike freedmen seized power and couldn't govern anything, and so it was a disaster in practice until white men got back in power and fixed everything. This view of Reconstruction is completely alien today, and it was made that way through decades of careful scholarship examining the era with new lenses, especially the lens of the freedmen as sophisticated political actors pursuing their own interests like any other political faction. The new view has largely overwritten the old view to the point where, speaking for myself at least, I was unaware the old view ever existed in the first place.

That overwriting process is happening constantly, and it's a direct result of the subjective nature of truth. What was true in the first 30 years of the century and what was true in the last 30 years of the century do not agree with each other, despite nothing new happening in the Reconstruction era itself, being frozen eternally in the past. Everything that will ever happen in the Reconstruction era had already happened, and yet the truth changed. And it remains possible, if not certain, that the future will have a different truth.

Umineko also has plenty to say about class, abuse, patriarchy, childhood development, the 1980s Japanese bubble economy, the mystery genre and its conventions, parasociality, the 72 great demons of the Ars Goetia, and the importance of investment liquidity. I confess that I have not often read written works of this length. One word count estimate places it at 1,154,971 words, or about 1.5 bibles. It's possible that any work this long has a broad list of topics to discuss, but to me it's the primary thing that makes Umineko worth the vast amount of time it demands (my reading took around 100 hours). There are stories that have nothing to say despite having many words, and this is the opposite of that.

During my readthrough, I had a discussion with a friend about how to tell people what's in Umineko to get them to give it a chance. I leaned on the subjective nature of truth, and she leaned on the murder mystery. What I wrote here is roughly what I would have written to myself to get me to play it, since all this sounds frankly incredible to me, but one must make up their own mind for such a commitment.

And you have to do it right now, because this is the end of the line for the curious-only audience. If you are still on the fence, please do not read the rest of this. You can never unlearn something interesting once it's been learned, so if you haven't yet read Umineko, then reading further would be an irreversible mistake. You can always come back and read it later, but you can never unread it. Take your decision seriously.

To The Enlightened

Presumably you have read the entirety of Umineko if you're here. If you haven't read it, but you feel like you want to read about it anyway because you "don't care" about spoilers or because you've sworn off the possibility of ever reading it, I don't think you would get much out of this. The events of the game are fresh for me, and I'll be assuming a familiarity with the story here, so a lot of this may not make sense. There is no plot summary here. I'm not the boss of you though!

The Ange And Maria Show Owns

First things first: my favorite part of Umineko by far was Maria's story in episode 4. I related to her, as a kid that was hungry for knowledge in a special-interesty way that often annoyed the adults around me, and the desire to be seen as an expert in anything. I obviously didn't have it as bad as her (my parents are lovely and my classmates were kind enough) but it was so easy to get inside her head and see things as she did.

Maria is the most on-the-nose demonstration of the subjective nature of truth, or how without love the truth cannot be seen. She was able to ward off depression and loneliness through the magic of her toys coming to life to keep her company, and she was able to attribute her mother's abusive episodes as the black witch possessing her real mom and taking over her behavior. Her happiness depended heavily on the strength of her belief, and she was the believing champion of the world for so long there. Ryukishi07 spent ten years working as a civil servant, and it seemed pretty clear that the scene with Rosa screaming at the social worker was likely built from the pieces acquired from that.

When Ange summons Maria by reading her journal, she says that Maria's situation is similar to her own: trapped in a school with people who hate her. We're presented with a dual timeline. In Ange's time, she's reading Maria's diary and seeing nothing but optimism recorded there. She's blown away by Maria's resilience and love. Ange wants that for herself. In Maria's time, she's actually struggling quite a bit as Rosa grows worse and worse, escalating to the point where she pledges to kill her. Nevertheless, Maria's fictitious words probably helped to give herself a mental health boost, and they also gave Ange quite real confidence when she needed it.

I think Ange's story has more to it than it appears at first. Ange's school has a system where everyone gets grades based on the class average in order to foster cooperation, and the group with the lowest average is made to volunteer for extra work. Ange is constantly failing because she is depressed and falling apart. Her classmates get together to form a study group, where they take their lunch and weekend time to get a study group together to help Ange get her grades up. Just like Maria portrayed a pretty grim scenario in a positive light, Ange portrays this pretty positive situation in a miserable light. To Ange, this is a group of bullies taking her hostage to berate her nonstop because they don't like her. But would bullies do that? Isn't taking all of your free time to get together with your class's worst performer a pretty time and labor intensive way to bully someone? Wouldn't it sort of make things way worse for their group average if they singled out a girl in their own group to harass?

Granted, they do a pretty bad job. They force her to write an insincere apology letter and further force her to recite it. I don't think they are very good at being benevolent tutors. I do think it's at least plausible that they were trying their best and failing, though, rather than trying their best at bullying and succeeding. Ange, at this point, is simply incapable of seeing the good side of anything, because she can't see it without love.

The Subjective Nature of Truth

Getting older and hearing more stories means that there are fewer and fewer gaps in the human experience to fill in; just about every random show can impart a life lesson to a teen, but in my 30s I think I have pretty much all of the basics ironed out and the only gaps left are details. Admittedly, I also thought this about myself when I was 16, so maybe when I'm 50 I'll think I'm a fool now, but that's where I'm at today.

If I'm taking anything life-changing away from Umineko, it's pretty much all from the story between Maria and Ange. Role-playing is sometimes a suspension of disbelief, but from within the frame of the belief it is completely real. If you inhabit the same frame, if you go along with the premises of a performance, you can exist in whatever way you imagine. There is no top level, beyond the reaches of role-playing, because even the "real" world is within some frame of a belief.

I recall the great Wallace Shawn talking about this concept¹:

Actors understand that we contain many, many people inside ourselves, and we could be a king or we could be a thief. Unfortunately, in the real world, people are cast as kings or thieves at birth, in a way, and they're not allowed to get out of it, but really, people are unconsciously faking, they're acting. They're assigned a role, let's say as a worker, and they eventually learn how to buy the clothes that fit their role. They're not even assigned those clothes, they actually learn how to buy them. They use the accent that other people in their role use, and they even pitch their intelligence to the level that they think is expected of them, all unconsciously.

If you get up on a stage, and you want to play a king or a queen, you may be startled to find that you can inhabit royalty to the extent that you really become royalty. Doing that enough, it could very well be the case that the things going on inside your head are essentially the same things going on in the heads of real kings. And while we're at it, what mechanism even determines who does and doesn't get to be a king? In Debt: The First 5000 Years, David Graeber has an answer:

Politics, after all, is the art of persuasion; the political is that dimension of social life in which things really do become true if enough people believe them. The problem is that in order to play the game effectively, one can never acknowledge this: it may be true that, if I could convince everyone in the world that I was the King of France, I would in fact become the King of France; but it would never work if I were to admit that this was the only basis of my claim. In this sense, politics is very similar to magic-one reason both politics and magic tend, just about everywhere, to be surrounded by a certain halo of fraud.

If you believe yourself a witch, and you act like a witch, and you answer any questions from the perspective of a witch, it becomes quite difficult for yourself or anyone else to prove that you're not. Even the concept of proof becomes fuzzy. Likewise, Ange wanted to prove Maria was unhappy, but there wasn't a single word in her diary that she could use as evidence. Who's to say she was unhappy? Ange tries and fails, and her failure ultimately inducts her into an uneasy acceptance of her own branch of magic.

When I was young, I was interested in magicians. I watched David Copperfield in particular, and I was fortunate enough to see him live once. I was totally enthralled by it, not yet having a full understanding of the parameters of the world and what things were and weren't possible. I basically believed that the magic tricks were being performed exactly as they appeared, and was not aware there was any illusion at all. I remember attending Sunday school later, where I was being taught about Jesus by an older lady who was describing Jesus's miracles. I said that they sounded like magic tricks, and she rejected this strongly; they definitely weren't magic tricks, they really happened. I said yes, right, they really happened! She couldn't understand my position, and I couldn't understand the distinction she was trying to make, and I eventually gave up. I remember it now as the point when I started to become suspicious of religion, but then again, I have re-remembered it so many times, it's impossible to say what the truth there is. In any case, I was fully ready to believe that both Jesus and David Copperfield were performing magic tricks. Who's to say they weren't?

Mysteries With Good Characters

Towards the end of Umineko, the character Will Wright² is talking about the goat demons, which at that point are reader stand-ins muttering theories as they attack. He says that not attempting to solve the mystery at all in a mystery work is the worst way to read it, and I felt some amount of guilt. In my defense, I did try to solve the rooms in the beginning, and I think I even got some of them right, but I didn't get anywhere close with the epitaph. I started to feel that the story as presented was so fantastical that any kind of "solution" would involve a random set of things presented in the story being illusions, and felt there wasn't enough foundation on which to really start any deduction. I would half-heartedly think of something, but once my pet theories were overturned by red truth, I simply didn't try very hard to overcome it, since I was not confident that they were solvable at all. One of the episodes even has a sizable section talking about this, where getting readers to attempt solving the mysteries requires a level of trust, which is the intended function of the Knox decalogue.

My shortcoming with mysteries feels like a problem of under-specification. When a solution to a mystery is revealed, my typical reaction is "well if I knew that this ambiguous statement meant that specific meaning, then I could have derived the solution myself". The red truth system actually goes a long way towards fixing this for me! In fact, in the early episodes I was at my most energized for solving the mystery part. I eventually faded out on that over time though, because the many twilights all started blending together. The level of word games going on in red ("humans" vs "people" having different meanings is just silly, come on) later reassured me that I had made the right choice. I greatly enjoyed reading Battler's theories and seeing what responses cut them down, and I can see the appeal in a riddle-solving process where you can go back and forth asking clarifying questions. I couldn't do that, but Battler could, so the consolation prize of seeing Battler do it was good enough for me.

Even in episode 8, in the most stripped-down logic puzzle form of the mystery solving process, there are frustrating ambiguities in the rules. Does "A culprit must not die" mean that the culprit of a particular scene doesn't die in that scene, or that they're still alive by the end? Does "From now on, Kanon is treated as being killed" satisfy the requirements to disqualify Kanon from being a culprit or not? Does "there are five master keys total. One is held by each of the five servants" mean that Shannon and Kanon have two? Can they use each other's? The problem here is that, given the golden logic path to the solution, these ambiguities don't actually end up mattering. But they definitely matter for ruling out non-solutions, which is the only real way to pin down the golden path in the first place. When I identify this kind of ambiguity, I immediately shut down and assume that I can't solve it, and I'll have to wait for the reveal to find out what the parameters of the problem even were in the first place. Don't get me wrong, I totally understand how this exact thing is incredibly fun for people, but I just gotta be able to collapse ambiguities to pursue a solution myself. Honestly, if I do put in a bunch of effort and solve something, the payoff I feel isn't "wow I'm really smart" but rather "the scene where the grand scheme is revealed is less interesting now". In Umineko, the epitaph is the only real riddle that has a direct explanation, anyway; the actual human solution to the locked rooms early on are just left as an exercise to the reader.

I think part of the situation is that I suspected Shannon early on. The very first time Beatrice shows up, Shannon was talking about how hungry she was and how she could smell delicious soup that she wasn't allowed to eat. Beatrice's first lines in this situation are an elaborate metaphor about soup. Unfortunately, because Shannon was in the first group to die, it seemed like it was unlikely. It wasn't until the episode 8 puzzle that I really started thinking about how, if I'd tried just a little harder, if I had known about the possibility of fake corpses and pretend deaths earlier, I could have gotten some of them figured out. The misdirects were just too much for me though, with Shannon and Kanon having different deaths and being casually kept distinct in red truths. I think speculating about split personalities at that point would feel like cheating the system, it would violate something that seemed so obvious that it didn't require stating it outright, like "actually everyone on the island is an alien species capable of spontaneous resurrection" or something.

The primary thing I read stories for is learning about characters, and how their past shapes who they become. There are many great stories along those lines in Umineko, especially in episodes 4 and 7. However, the mystery genre presents an obstacle to my type of reading because it requires that all character development and motivations be potentially suspect. I noticed that a lot of Battler's theories required people to act for no reason whatsoever; people doing tricks, faking their death, reanimating at specific times, going to specific places for no reason at all besides acting as marionettes hitting the checklist of events that a particular logic puzzle requires to happen. And indeed, it's later suggested that Beatrice was able to do these things through the power of bribing accomplices gold, so there would be no clues in any character's personalities besides that they all need money really badly.

That said, not every character is given the same level of development. The children (Battler, Jessica, George, Maria), the wives (Natsuhi, Eva, Rosa, Kyrie), and the young servants (Shannon, Kanon) were all super fleshed out; by the end, I feel as though I can tell you where each one of them are coming from, why they're doing what they're doing, how their behavior sprouts out of their upbringings, and so on. I'd put Kinzo in this group too. I cannot claim to be able to do this with the husbands (Hideyoshi, Rudolf, Krauss), the other servants (Kumasawa, Gohda, Genji), or Nanjo. Oftentimes, these characters felt like their only personality was individual traits (mackerel jokes, foodie, serious professional, etc.) at best, with little history. They tried giving Genji a little history, but it really didn't paint a clear picture of his development. In this sense, it would be easy to suspect these as the culprits, but within their thin characters they all have reasons to be allegiant to other characters or the family in general.

I understand this is just a fact of life in the mystery genre, that someone involved has a hidden backstory and is presenting a false face to the rest of them, but this is totally anathema to how I'm reading stories and internalizing information about these people. There were clues that some deception was going on, of course, but it's something that I actively didn't want to believe, so I was motivated to ignore them. To fully participate in the mystery, I would have had to look at things like the love scenes between Shannon and George and go "but what if none of this is real and it's all an act?". My mind simply resists this, because if I'm not getting an earnest account of a character's disposition, then the whole work feels cheap. I did enjoy the later parts where some of the deception was revealed along with the in-character reasoning for doing so, and I liked looking back to see how the murders could have been done with this new information, but I have to say in response to Will's statements that "never disregard the heart in mysteries" and "always try to solve the cases as a reader" are contradictory to me. I'll take a compelling narrative with a mystery to move the plot along (shoutout to Shoushimin for being this), and I'll take a stripped-down word puzzle where the characters are all flimsy paper pawns (as in Bern's episode 8 puzzle), but it's simply too hard for me to try to guess which one a given story is when it could go either way.

Fantasy And Mystery Authorship

In episode 7, Gaap presents a theory of mystery and fantasy:

Very entertaining, these locked‐room murders. Once you read the answer, it becomes a tale about humans, or a mystery novel. But, if you rip out the part that reveals the answer and throw it away, it becomes a tale about witches, or a fantasy novel.

In other words, a fantasy story is one where you simply take the detective's explanation at the end and cut it off, leaving the whodunit and howdunit unexplained, thus allowing you to believe that the crime was carried out with magic. I think that applies to the specific mysteries in Umineko, but I have some problems with that as a generalized principle. In mysteries, the whole point is that everything is written with invisible logic behind the scenes. In the author's head and/or notes, there are real mechanisms that sharply restrict the possibility space for what can happen. The process of writing mysteries (or being a game master in Umineko) requires the author to carefully set up every crime to make sure there's a path remaining for their human culprit to carry out the crime without creating any logical inconsistencies, and then to obscure that path as much as possible through misdirection and misleading clues. If we're simply chopping off the end of a mystery, then the resulting "fantasy" story will still have all of that careful attention to detail baked in. The implication, though, is that fantasy written as fantasy from the start can have anything happen for any reason. The fantasy author is free from the constraints of logic and can summon anyone, do anything, and abandon cause and effect entirely. They can create a totally baseless spectacle.

Fantasy still has to follow its own internal logic to be any good, it's just that the logic is based on the themes and emotions of the story rather than carefully constructed alibis for all of the characters. It's true that you can simply have a wizard cast a spell winning the day; there's no limit to what words you can write on a page. However, actually doing this is the fantasy equivalent of saying that the killer has the death note or some other decalogue-violating implement. When a wizard casts a spell, it has to correspond to how the wizard feels in the moment, how strong their position in the story is, whether it would be emotionally resonant for their opponent to be struck or for them to somehow evade the blow, and so on. The reader's role is very much the same as with mystery: discovering the themes of the story through the semi-obscured rules governing it.

I'm a theme fan. I love when a work expounds upon some themes and studies some characters. And so I despise, a little bit, the idea that fantasy represents a kind of narrative surrender. I get lots of enjoyment out of solving characters, instead. As it's put in Episode 7:

People are riddles. They want someone else to solve their riddle. They live life wanting someone to solve the riddle that they are, the most difficult riddle in the world. They want someone to look at the riddle they are. And they want that person to solve it.

I love this passage for a lot of reasons, but mostly because it affirms the way I was reading the story up to that point. Solving the epitaph would have been pretty satisfying, but I feel like solving how Ange's and Maria's circumstances shaped the people they became gives me a much richer bounty that I can carry with me long after I finished my reading. In fact, getting good at deciphering this kind of character riddle is one of the things that gives meaning to being alive. If I couldn't do this with media then I would probably not bother with most of it.

Being Good at Mysteries

I want to confess here that I'm open to the possibility that this is all just cope. Maybe I got a bad grade in mystery and so I'm scurrying to the safety of fantasy to shore up my bruised ego. Well, what does it mean to be "good" at mysteries? I think Umineko offers a answer to this with Erika, the "intellectual rapist". When she first showed up, I was hoping that she was a sort of facts-and-logic debate type person that would shut down the witch illusion with rationality, and she's not not that, but she has some extra qualities tacked onto her character beyond that. Despite insisting that she is always seeking the truth, there are several instances (particularly the Natsuhi trial) where she'll happily play defense around a cohesive, internally consistent narrative even in the face of evidence that it probably isn't the truth. She uses the whole exercise as a stepping stone to get to her real goal, praise and recognition.

There are people like this, out in the world. A common mistake in identity formation is to place all of your self-worth in the single basket of being "the smart one" (as Erika is), or "the creative one", or "the athletic one", or some other singular category of excellence. The problem is that when you have no other pillars holding up your identity, any threat to the one you have is a personal threat to who you are as a person. When Erika's theories are challenged, it's not just about whether someone could leave a sealed room or not, it's a personal attack claiming that she has nothing remarkable about her at all and she has no reason to exist. Of course, when those are the stakes, it's quickly going to escalate into emotional shouting matches and denials. When you can't move to a different pillar to stand and look out at the situation with some distance, you can only see the collapse and the ground rushing up at you, and the only way to save yourself is to talk the pillar back into existing before you hit the bottom. Before you know it, you're shouting at a nine-year-old girl that she actually got candy through a fake illusion and all her dreams are bullshit. Terrible way to go.

In order to get good at mysteries, it also seems like a requirement to gain fluency in the genre as a whole. There are a finite number of patterns that murder investigations can take, and when you're able to categorize them and understand how every type of investigation works, you're starting from a point much closer to the solution than someone who just started out. It's not unlike Maria's wolves and sheep puzzles; after doing them repeatedly, you can settle into strategies and algorithms to make the answers appear much faster than someone trying to solve their first one. I think any task like this can eventually become easy with enough practice, and that the marker of aptitude is not how good you eventually get, but rather how little practice it took to get there. Erika, being purpose-built for solving mysteries with thousands of years to absorb every work in the genre, is not particularly impressive in this regard!

I am a software developer by trade (though hopefully not by identity). In my experience in the field, there exists a constant temptation to view other people as lesser. Oftentimes people who do software engineering believe that, because they have mastered one complicated domain, and that they are applying their skill to automate some small fraction of someone else's job, that they are actually capable of all jobs simultaneously, and that outsiders are only capable of one. This is catnip to a fragile ego, to the type of person with a singular identity from which they derive all self-worth. I have found it important to take special notice when I'm bad at something and give it more weight than I reflexively feel like I should, in order to keep this feeling at bay. The rigorous tools of programming have a limited domain. Not every problem is accessible to objective solution; joy cannot be examined in a microscope, love cannot be the product of a formula, and consciousness cannot be reduced to chemical reactions. Trying to do so leaves the majority of the human experience abandoned on the table, which is why so many STEM type people come out as sociopaths. They never developed the different tools required to interface with the soft questions of humanity.

Otaku Bait

Speaking of soft questions, I'd like to explore the permissibility of incest. Kinzo had four children with his unnamed arranged wife, one of whom fathered Battler. He later had a child with Beatrice, and then with his daughter, to make Sayo Yasuda into Kinzo's half-daughter and half-granddaughter. Umineko is, among other things, a love story between Battler and his father's half-sister/aunt. I found it kind of surprising that Kinzo's familial bond with the second Beatrice was used to paint that relationship as disgusting even above and beyond the non-consensual nature of it, while Battler's familial relation (consented to, hard to overemphasize this, very enthusiastically) is totally unremarked upon by anyone at all. George and Jessica are both equally as related to their respective partners too, for that matter. I suppose the Ushiromiya family has bigger fish to fry than to worry about that, once anyone knows that's what's going on.

Casual incest is not as common as it used to be in modern anime, along with other prior staples of the medium. There are quite a few 90s/00s mainstays in the opening of the first episode, with the groping and upskirts and so on. I wonder about what the motivation for this was, since the rest of the game doesn't keep that tone at all. I think the idea is to lure in otaku who are used to playing bishoujo games and the like, to set up expectations for a flirty vacation time in order to make the horror and murder extra shocking. Unfortunately, for people not forged in the fires of this stuff, it has scared a lot of people off the VN entirely.

There's a wrinkle in this though: eventually, it's revealed that the first episode is not actually an account of events happening in-universe, but rather a story written by Beatrice, huge mystery fan that she is. So Beatrice is writing her fanfic about Rokkenjima and the Ushiromiya family, and she decides to write a scene where Battler starts to grope Shannon's chest but hesitates when it looks like he'll actually get to do so and then gets punched by Jessica. When Beatrice sits down with a pen and cooks up a story, this is the scene that sprouts from her imagination. The stories are, after all, repeatedly stated to be a kind of love letter from her to Battler to convey her message to him, so I suppose anime hijinks are part of that love.

Meta-Fiction Models

I have stumbled onto a rather sticky topic though, which is the meta-fiction situation. Events are depicted directly which didn't happen, but rather were an in-universe story written by a character in a higher-level story. So let me ask a question that is far more difficult than it sounds: what actually happens in Umineko? I've written plenty on what Umineko is about, and I feel like I have a good grasp on the actual events that took place on the island on October 6, 1986 and the years leading up to it, and I'm surprisingly clear on the timeline of events in the world at large. But I want to ask not what happened on Rokkenjima, but what happens in Umineko, the visual novel.

The Secondary World

I am going to discuss this at some length, starting from a foundation and building up ideas about both fiction in general and Umineko specifically. Let me start with the most basic question: what are you reading when you're reading a story? If I write a short story that says "Alice shot a fireball at Bob", what's going on? Well, I'm conveying to you that there is a world that exists. I, as the author, know the events of that world. I am reporting those events to you. If my story is written well, then you will perform literary belief and imagine the world I'm describing. J.R.R. Tolkien³ gives us a useful framework for this:

Children are capable, of course, of literary belief, when the story-maker's art is good enough to produce it. That state of mind has been called 'willing suspension of disbelief'. But this does not seem to me a good description of what happens. What really happens is that the story-maker proves a successful 'sub-creator'. He makes a Secondary World which your mind can enter. Inside it, what he relates is 'true': it accords with the laws of that world. You therefore believe it, while you are, as it were, inside. The moment disbelief arises, the spell is broken; the magic, or rather art, has failed. You are then out in the Primary World again, looking at the little abortive Secondary World from outside.

I am going to use this secondary world terminology quite a bit, and I need to define some terms here to make this readable at all. The primary world described by Tolkien here is, of course, the real world. Ryukishi07 and I are both in the primary world, and so are you. Inside the visual novel's world is the world of Battler, the Ushiromiya Family, and so on. When I talk about a character in the secondary world (that is, the uppermost layer within the fictional universe), I will use nomenclature like "2W Battler", meaning "the Battler in the secondary world". When we first meet Battler, though, we actually are not seeing 2W Battler.

The Layered Authorship Model

The first episode is a message in a bottle that 2W Beatrice wrote sometime before the murders and cast off into the sea. This is a story within a story, a tertiary world. The first Battler we meet, then, is actually 3W Battler. The first impressions of all of the characters are tertiary visions of them through the lens of 2W Beatrice's authorship. This is totally fine, and I actually like this setup quite a bit; I loved seeing later on how real-world elements influenced the authorship process.

I also enjoy thinking about the character perspective through which a 3W character can ascend to 2W. When I say "2W Beatrice", I'm using it as shorthand for something like "2W Sayo Yasuda's body, which is being driven at that moment by the Beatrice personality that she has manifested". When 2W Sayo Yasuda is having a truly miserable time at the mansion, between Natsuhi's constant pressure, the bullying of the other maids, and Battler's forgotten promise to her, she recedes into the tertiary world of fiction. She obsesses over stories and she begins to develop hallucinations of different personalities. When she first encounters "Beatrice", she actually has Gaap's appearance. This is a tertiary world character at this point. That's followed by a process of revision for this character: she creates a witch in white (to oppose Gaap's red), who hates mirrors (she hates looking at herself), and the motifs of butterflies and spiderwebs from doing free association with the objects around her, essentially. That character becomes Clair, a kind of beta Beatrice, which undergoes further revisions with the western-style dress from the portrait and Battler's taste in women. Once the character is complete, it exists so strongly in her mind that she can perform the feat of being "possessed" by her, where she acts entirely as though she is the character. At this point, it ascends from the tertiary world of Sayo Yasuda's fictional universe to manifest in the secondary world alongside her. Earlier, I quoted Wallace Shawn discussing how actors in the role of kings can come to inhabit the very same headspace as kings, and how simply living your "real" life is a performance; it's the same here. Beatrice "exists" insofar as she is who Sayo Yasuda is deciding to be in those moments.

So to review, in Umineko we are presented with stories that exist at several different world-levels. Sometimes, we are presented with a depiction of the secondary world. Sometimes, the tertiary world. Sometimes, the secondary world with one character's tertiary fantasies inserted seamlessly, with its disentanglement being an exercise for the reader. Let's try walking through a reading with this layered world system in mind and see where it gets us.

As previously mentioned, the first episode is 2W Beatrice's story with all of the characters being 3W versions. The first tea party self-references Umineko itself, so I think it's safe to say that at least that section is a kind of non-canon author-insert type deal, which I'll happily excuse from scrutiny. The first ??? chapter introduces our first problem with this system, because it mentions Bernkastel. What world does Bernkastel exist in? Well, we eventually learn that Bernkastel is based on Hachijo Ikuko's cat. 2W Beatrice, the author of the first and second episodes, does not know about this cat, obviously. It was almost certainly not even born when she was writing her message bottles! But let's allow an affordance here and say this, too, is a non-canon extracurricular thing, since it's stuffed away in the ??? section.

The second episode introduces several meta-fictional elements. We learn about the concept of the game board, Beatrice using the red truth to give hints and goad Battler into trying to solve the mysteries. Remember, the direct events on Rokkenjima are all taking place in a tertiary world already, penned by 2W Beatrice. Battler and Beatrice's arguments in the separate, meta-world space must be different, then, since they are quite clearly spectators and can't interact directly with the other 3W characters. The only options left, then, are for these arguments to be happening either in a parallel 3W, which must belong to 2W Beatrice (as nobody else knows about the messages yet to be in a position to write or imagine the conversations), or a 4W where one of the 3W characters is imagining them.

I don't believe that these arguments are actually in the text of the message bottle. None of the witch hunters ever mention the meta-fictional elements of the red truths or Battler's arguments, which one would think would be a central concern for a group of people trying for years to decode the mysteries. I think the content of the in-universe message bottle text simply states outright that magic was used with no mitigating explanations at all, aside from 3W Beatrice perhaps bragging about it. Therefore, the only real possibility is that these meta-fictional debate scenes are 2W Beatrice's imagination, daydreaming about her beloved Battler participating in her game with her as she writes the stories.

So, I actually like this! I think this gives a tremendous insight into 2W Beatrice's character, her inner life, motivations, and so on. This insight comes at a cost though, which is that 2W Battler has not appeared in the story yet, and we know essentially nothing about him. We can infer a bit about him through 2W Beatrice's impressions, but it should be pretty clear that meeting someone at age 13 and obsessing about them for years does not create a particularly representative picture. As I mentioned in the previous sections, I am a character fiend and I love learning about characters, and I would even love learning about 3W Battler, a character in his own right, but let's continue and see where that leads.

In episode 3, we leave this 3W Battler behind and we never see him again. This is because Hachijo (either Ikuko or the Ikuko/Tohya pair) begins writing forgeries, and authored the next few episodes. To distinguish these two versions of Battler, I'll call 2W Beatrice's character 3WB Battler, and Hachijo Ikuko's character 3WH Battler. Hachijo's character is also a reflection of 2W Battler, because Tohya has his memories and thus can pretty thoroughly explain the kind of person he was to Ikuko, but Ikuko doesn't have any connection to 2W Beatrice besides being a reader of her stories. Nevertheless, 3WH Battler seems to continue his character development seamlessly from 3WB Battler, building on the red truth arguments from 2W Beatrice's unwritten daydreams. 3WH Beatrice seems to have the very same affection for Battler, and continues saying the game board is a message from her to Battler which she desperately wants him to solve; this love is no longer originating from 2W Beatrice though, but rather 2W Hachijo making a 3WH Beatrice to simulate what she and Tohya know of 2W Beatrice, which is frankly not much.

At this point, our fiction-within-fiction model is in pretty bad shape. It is dealt a fatal blow in the opening of episode 4. Ange is on the rooftop in 1998 and Bernkastel intervenes to whisk her off to 1986 to help Battler. This cannot be in Ange's head, because Bernkastel is Hachijo Ikuko's character, and Ange and Ikuko have never met. This cannot be in Ikuko's head, because she does not know Ange's situation at all at this point with this level of specificity. She would have read tabloid articles and the like, but the Ange storyline here culminates in Ange meeting Ikuko in the real world, starting from this rooftop encounter. There is a chance that the entire world of 1998 as depicted is invented in retrospect by Ikuko, after meeting Ange decades later and imagining the trials and tribulations she went through to reach her, at which point we'd shift everything down a level (2W Battler becomes 3WH Battler) and the only straightforward secondary world event that is actually happening as described is the meeting at the very end and the visit to Gospel House. But at that point, the entirety of the story gets flattened into Hachijo's pocket universe, so in a sense every character just becomes a single flat 3WH version, which somewhat undermines the whole thing. This a fairly extreme a shift in the nature of the story to read into such small details.

So, without doing a play-by-play of every time this layered world model breaks in the game, interpreting the story strictly as layered fiction makes it worse on average. If this is the reality of the situation, then we learn almost nothing about 2W Battler besides the 1980 promise, the family conflict, the 1986 return, the post-incident escape with Beatrice, the boating accident, and the traffic accident, at which point he "dies" and becomes sealed up in Tohya. In this layered fiction model, information is scarce about any of it.

I have a very important caveat to this criticism though, which is that this did not occur to me whatsoever until I sat down and started writing this essay. Once I saw the magic ending with Battler and Beatrice drowning, I started thinking about the episode 7 depiction of the "real" events of the island to think about which elements had to change to support this boat escape. Beatrice appeared to be killed, but must have avoided it (as Eva did), so maybe Battler wandered down the secret passage opening behind the Chapel to find Beatrice there. What I badly want to know is: what on earth was that conversation? Did Battler remember the promise that Beatrice has been obsessed with? Did she tell him? Is she upset he didn't send Shannon a letter when he sent one to his siblings? We simply can't know, it's sealed in the cat box. From that seed of curiosity, I realized how vanishingly little we knew about 2W Battler. I started writing about the layers involved, which characters belong to which writers, and realized how the system breaks apart on the shores of Bern.

I am of the opinion that if you're only bothered by something after you finish a work, it's not a real problem with the work. The only job Umineko had was to entertain me while I was playing it, and good lord it did that well. If the illusion crumbles on reflection, that's fine, and even a little fun. Who doesn't love being a pedant spotting a mistake? So even though this section might sound critical, I want to say here that it was at no point a negative for my reading experience.

The Fantasy Model

There is an escape hatch from this problem, though. The escape hatch is that magic, seemingly constrained to the tertiary worlds created by the characters, is actually literally real inside the secondary world of Umineko. Battler is wrong. Beatrice is right. No human solutions to the puzzles are required, she really did all that with magic, and really did create a game board that makes the events happen repeatedly. Bernkastel literally manifested to Ange in 1998, and pulled Lion from another timeline, and Featherine really did have Bern as a miko and named her human alter-ego's cat after the witch herself. The SSVD and the Witch Senate and the City of Books are all literally true in a kind of hidden civilization kept secret from ordinary people. Could this work? Sadly, this runs a little counter to the overall message of the game.

One piece of evidence in the text for this interpretation is in episode 6, when Beatrice is recreated from scratch but does not have any of the traits of the original. There are two Beatrices, an older and younger sister, each talking to each other and synthesizing their different personalities to try to recreate the original thousand-year-old witch. This seems (only in hindsight) like a direct acknowledgement of this problem, with the older sister being the 3WB Beatrice character from the message bottles, and the younger sister being the creation of Tohya in reality, and by Battler the Game Master in the fiction. They are trying to merge into a single character that represents 2W Beatrice. This thematically aligns with Shannon, Kanon, and Beatrice competing for who gets ultimate control of their shared body to commit to their romantic partner. Unfortunately, it's not at all clear who is making this tertiary world; this is a situation that cannot happen in the secondary world (Jessica and George cannot see Shannon, Kanon, and Beatrice in the same room), and so someone is imagining it, but I have no clue who, or when, or where. Another mystery, I suppose.

I wrote earlier about the fantasy/mystery dichotomy, and how the game's idea of fantasy is a story where literally anything can happen without any internal framework or logic, and I said that the framework is about themes and feelings rather than strict adherence to literal cause and effect. I think this theme-forward view is the only frame in which the game can make sense. In episode 4, Shannon and Kanon are imprisoned in a jail cell with the rest of the family, meaning it's required to be a fantasy scene. At the same time, the cousins are called out to do tests, and Battler's test is hearing Beatrice talk about his sin six years prior for the first time, and to try to remember what it was. The prison cannot be real, but the test cannot be fake, either; this is a crucial turning point for Battler's character, with threads reaching all throughout every version of Battler that appears. It is thematically coherent but factually incoherent, just like fantasy ought to be.

So, is there a better model for thinking about the way the multi-authored characters exist together in the story of Umineko? The way the story is written makes it fairly natural to assume that there is a single Battler character, and he has a kind of psychic cloud save support. Every writer in every era can download that single instance of Battler with his own preserved character progression, relationship status, foreshadowing, and so on. Any writer can write their own story, stay faithful to everything anyone wrote about him, and put him back in the cloud for the next writer.

The Grand Narrative Model

There is a literary concept called "narrative consumption". I have not read the major works, but I did read an essay related to the argument. The basic idea of narrative consumption is that there is a main world, or worldview, that forms the basis for a story. Derivative works (for our purposes here, the message bottles and forgeries) draw on this central narrative, and put their own twist on it. This pattern shows up in both kabuki and doujinshi works. The kabuki version of "world" is defined this way⁴:

A concept that refers to the historical era or events that constitute the background of the work. In fact this concept includes everything from the names of the characters that appear in the work to the basic personality traits of these characters, the nature of their relations, the basic storyline, the basic aspects and developments that should be dramatized, and so on. While this "world" is mostly founded on the commonly known popular history of Japan, oral traditions, and so on, it also contains generic content developed through the repeated dramatic adaptations and performances in the form of preceding Kabuki and Japanese puppet theatre as well as medieval performing arts, and thus it does not necessarily refer to any established sourcebook or original text. Therefore, each individual "world" is not a permanent or unchanging thing; some new "worlds" emerge and others fall into disuse and remain in name alone as a result of the formation of genres as well as the fashions of the time. The authors thus create their works by dramatizing newly invented "variations" that are based on a particular world, which is commonly known to the actors and their audiences, or by mixing multiple "worlds."

I think this precisely describes the Umineko universe. There is a shared fabric of the story, the Rokkenjima Serial Witch Murders, with much of the detail shrouded in the mystery of the cat box. Multiple authors are taking that base premise and developing it into derivative works. Some of them read each others' works and incorporate them into their own understanding of the main base story, and have them all swim together in a big soup of meaning. This doesn't totally fix the problem of Bernkastel's dependency on Bern the cat, but it gets us most of the way there. We are reading, effectively, a series of doujinshi works about the Rokkenjima incident. Those stories are potent enough to spawn their own derivative works (the meta-fiction conversations), which spawn yet more meta-meta-fiction works and so on, into an unbounded number of layers. There is perhaps a hierarchy of some sort here, but it looks more like a network than a flowchart. All of the works influence all of the other works, and are influenced by them in turn.

This changes the nature of how the story is presented quite a bit. What Ryukishi07 up in the primary world has done for us is to curate a particular series of in-universe fanworks that chart a satisfying path for Battler and Beatrice, out of the shared morphogenic field that everyone is pulling from. It's not necessarily that all of these tertiary-world works were directly written with the previous and later works in mind, but rather that the collective arc that they formed was manufactured through the selection process. What Ryukishi did as the author, in a literal sense, is sit down and plan all of this out and write the words, of course. But he also has pulled us into a secondary world in which near-limitless derivative works about the famous Rokkenjima incident exist, and from those works created a kind of "watch order" out of them. Where one story leaves off, he scours around this theoretical in-universe pile of Rokkenjima doujinshi to find one that seems to pick up more or less in the same spot, and contributes its own development to the shared characters. When that one ends, he repeats the process and finds the next one, and the next one.

When Bernkastel talks about sifting through "fragments", it could be the case that she is scrolling through thousands of fanworks, filtering on tags, and trying to find one where Lion exists, or where the family comes home, or where the epitaph is solved, or some other outcome that was theorized strongly enough for someone to write a fanwork about it. The "Eva culprit theory" and "Rudolf/Kyrie culprit theory" are fan movements that catch some virality and end up forming their own cluster of works. This is how Beatrice of 1986 can drop the episode 5 line "Ayatsuji Yukito doesn't make his debut until next year". Many problems work themselves out under this model.

That said, I don't think this is perfect; I'm sure it falls apart in some instances. It's just a coincidence that Hachijo and Beatrice wrote all(?) of the works worth presenting? Out of all of my thoughts about Umineko, the layered-fiction part is the one that I am going to spend the most time thinking about as the game settles more and I get further removed from the specific details and only remember the broad strokes. I've rewritten this section more than any other as I kept developing the thought and I doubt I've finished the process.

Highlights

That was an exhausting and heady diversion, so to bring this back on track, I want to just talk about a bunch of individual scenes I really resonated with that didn't fit into any of the points I made so far.

George the Woman Respecter

In the section about 00s anime staples, I talked about how rowdy and perverted Battler is early on. As a good contrast, I quite enjoyed the scene where George is describing how he fell in love with Shannon. He used to feel a huge amount of caution around women for fear of coming off as "one of the bad ones", to the point where he was no longer treating them as human. Then he saw Battler making crude jokes and being the life of the party, violating all the rules he made for himself that only he has to follow. Instead of suffering for it, Battler was thriving! George figured that the moment he did any of that stuff he was going to suffer complete social rejection, but he had a role model here that was doing just fine. That let him drop his guard a bit and have fuller interactions with women, including with Shannon, culminating with them forming feelings for each other⁵. I think this happens often with men. Raging misogyny gets a lot of social attention, as it should, because it's the type that actually leads to serious problems. George's flavor of benign chivalry has its own costs though. A world of pre-change Georges would be far from ideal. It was a pretty short scene but it packed in a good description of this nuanced social dynamic that I haven't seen too many other works attempt to address.

Battler's Test

The episode 4 test of Battler was fantastic as well, leading into the finale. The end of the question arcs really slammed in a lot of questions right at the end, the most important of which was "Who am I?". Throughout the question arcs, while I was puzzling over the red truths with some effort, one gap I couldn't figure out was what exactly witches were in the context of red truths. When Beatrice talks about the number of humans on the island, is Beatrice not a human? Are Gaap, Ronove, the Seven Demon Sisters, and so on? How about self-described furniture like Shannon and Kanon? Certainly they would say they aren't human, and in fact they spend a great deal of the story explicitly saying "we're not human". Beatrice says this, all in red:

Ushiromiya Battler. I will now kill you. And right now, there is no one other than you on this island. The only one alive on this island is you. Nothing outside the island can interfere. You are all alone on this island. And of course, I am not you. Yet I am here, now, and am about to kill you

I didn't have any idea how all of this could be true together, and reading it back now, I still don't. It's not hallucination ("I am not you"), it's not authorship (that's "outside the island" surely?). Maybe a trick of saying no "one" because Beatrice isn't "one"? It feels nonsensical now, but at the time, my brain was totally alight from this scene. I loved the rapid-fire blue and red, and I loved that Beatrice was accepting answers that were ridiculous even if they weren't the actual solution. I'm still holding out for the small bombs theory! It feels great in hindsight because, even though the red seems wrong, what this scene represents is a mystery novel fan finally getting to talk about her shared interest with her beloved and having it just pour out. Truly, nothing in the world feels as good as this.

Logic Error

The other time I was super motivated to work on the mystery was in episode 6. The fantasy characters were pressing the reader quite hard to try to work out solutions to Battler's locked room scenario, giving me lots of time and narration to roll the problem over in my head. I only got as far as something with the seals in the guest house, which is what ultimately ended up solving it, but in a more specific way than I had organized. It was fun to examine the situation from every angle and strain as hard as it felt like the characters were straining, and it was satisfying to actually get a plain explanation of the human solution, which the earlier twilights never got⁶ Honestly, episode 8's interactive riddles (especially Bern's puzzle) felt like the author saying "oh, people are really not doing any of this at all, huh?", and giving a late explicit tutorial for the concept of solving mysteries. The rest of the puzzles at the party were a little bit tedious, but I really liked the puzzle where George explained the Monty Hall problem to six-year-old Ange. Truly, this is the height of fantasy. Many grown adults live their entire lives without being ready to have the Monty Hall problem explained to them, so Ange is awfully precocious for picking it up.

The Salo Republic

One of the biggest surprises for me was in episode 7 when Beatrice Castiglioni⁷ stepped off the sub. It was so satisfying to see all of the elements click into place like dominoes falling: that's why he loves Beatrice, that's why there's so much emphasis on resurrection (because Kinzo says he felt like he was essentially dead all those years), that's what the gold is, the underground tunnels, the crater, so much snapped into focus through this section. After such a long work, having a flurry of answers to the mysterious elements was so satisfying. I have seen a great number of puzzle box media try to make an interesting setup and then just keep setting up until it runs out of steam and can't pay any of it off, so it was honestly pretty shocking that so much of the complexity proved to be under control in this sequence. It also had some of the best romantic chemistry on display in the game; George and Shannon were pretty good, Bern/Lambda and Battler/Beatrice are on some level beyond chemistry, but Kinzo and Beatrice had such a good rapport. They have great personal reasons that drove them into each other's arms through their shared history as the cast-offs of noble families. I was eating good. It's too bad that Beatrice is technically a fascist party official here but the story doesn't spend too much time working that out so neither shall I.

Lambdadelta's Identity

Speaking of Bern and Lambda, once it was revealed that Ikuko's cat was named Bern, I tried re-evaluating the whole game in this new context. Ronove is Genji, Virgilia is Kumasawa...was Lambdadelta based on anything? If so, what? The best thing I came up with was that she's a cat toy. Think about how cats play with their toys: they stalk and attack them, rip them to shreds, but can drop the hostility on a dime and switch to nurturing instead. The object of affection is prey rather than equal. That's exactly their dynamic! There's gotta be like a little stuffed candy plushie knocking around the Hachijo household. Maybe her name comes from needing a third witch to complete the dynamic with Beatrice and Bernkastel, and her name ΛΔ represents that third position being filled in. Love this kind of thing.

Beatrice's Contrition

Rewinding a bit to episode 3, I was totally taken in by Beatrice's performance of contrition. The exchanges where Battler started essentially protesting the game were especially good. So often, there's someone who is engaging in some kind of bad behavior⁸ and doesn't even realize what they're doing. A friend or victim will have a final straw broken and release an outburst about how they've had it and the perpetrator needs to stop. They give an ultimatum about it. The perpetrator feels it's completely unreasonable and out of the blue. If it's this serious and gotten this bad, why were there no signs before? The longer it's tolerated, the worse it gets.

Beatrice goes through all the phases: ignoring it, laughing it off, trying simple apologies, complaining about the lack of warning, and eventually going through anger and sadness about the loss of a friend for a reason that, to her, feels trivial and temporary. In her view, this is a big misunderstanding, and if only they reacted the way they'd been reacting before, everything would just go back to normal. Battler does exactly the right thing here: sustaining his view while giving her space away from him. If he applied too much constant pressure, the principles would calcify and the defensiveness would become reflexive. No introspection would be taking place, it's just a pure matter of the prompt fetching the response and terminating the thought.

This ended up (allegedly) being all an act, and Beatrice has some advantages in her performance. She's an excellent actress, as she says, of course, but she would be under much greater pressure if Battler was clumsier at trying to reform her. If he was screaming at her 24/7, well, there's not much opening for her to go "oh okay I agree now" on the 50th request while feigning sincerity, even if it was ostensibly what Battler would want to hear. It's also the point in the narrative where a redemption arc would actually insert itself! I was sold on it in large part because, if there was going to be a redemption arc, this is the exact place it would go, so any suspicion was melted away by the narrative position it appeared in and the strength of the writing. I especially liked the scene where she momentarily resurrected Shannon at George's urging. The narration describes her retrieving Shannon:

Those words of power knocked forcefully on the door to the Land of the Dead. The magic of her old self had performed that same forceful knock. However, doing it that way was horribly violent... and earned the displeasure of the peacefully sleeping dead... But this time was different. It was very forceful... and yet gentle and loving.

The dead all work together to find Shannon and deliver the casket to Beatrice, and ferry her back to the land of the living. "It was solemn, but filled with joy. The long-forgotten dead were made to remember that they were loved after their deaths, and they were all given a warm sense of peace." This is true evidence of reform, to me. Those that had spurned her were giving her another chance, which she's using to do a true kindness for someone else. It also requires some love! Without love, the people who she had wronged would continue to see her as dishonest and reject her. With love, they're capable of seeing the good in her and accept her attempt at contrition. By the end, she turns back into the cruel version, but I was never convinced that this was all an act. It seemed more like this was revealing to her true nature, and that the return to cruelty was at least partly a performance.

Conclusion

I liked just about everything Umineko had to offer while I was reading it. The only complaint I really had during the reading process itself is that the early pacing can get a little dull. Going back to the start of the mansion each episode knowing the situation is reset and there's going to be a new set of murders made me dread starting episode 3, since I understood the format at that point and didn't yet know the variety that it was going to bring with the redemption pivot. This kind of tedium ultimately led to me abandoning my attempt to read Higurashi, which I fell off of in the middle of episode 3. Knowing that episode 3 in particular is where the doldrums are helped give me motivation to continue and see the other side and I'm so glad that I did.

I've found that I have an affection for even just the concept of a long and acclaimed work. I like having the identity of "person who has completed Umineko". I like the thrill of wide-open expectation at the beginning and the satisfaction of conclusion at the end. I get this sense of grand adventure from several places like One Piece, Final Fantasy, and Yakuza. Umineko is definitely up there with those in sheer scale of time and emotional investment. Closing the large iron door on a story like that gives a sense of accomplishment that can't be matched elsewhere.

The main thing I'll be taking away from Umineko is its most repeated line: that without love, it cannot be seen. On the wiki page for Sayo Yasuda, it lists her likes as mystery novels and her dislikes as herself. I've spent a lot of time thinking about Sayo's journey and how it ultimately led to the murders, and I think the single most important trait is that she had no love for herself and therefore couldn't see her own value. Battler's promise becoming a load-bearing pillar of her identity was never Battler's fault, but it's hard to say it's Sayo's, either. She was never taught how to receive love, which has a surprisingly steep learning curve. Her life was punctuated by dismissal, between being removed from her mother, passed off by Kinzo, dropped off a cliff by Natsuhi's servant, raised in an orphanage, and abused as a servant herself. We know what would have happened to her if she had been raised with love: she'd be Lion. Under no circumstances would Lion need Battler's affection to not carry out serial murder, they⁸ have enough self-confidence to let that pillar go and rest on the others.

Maria and Beatrice are best friends. A part of Beatrice hates herself, and a part of Maria's true magic is loving herself. Ange had six years of love and twelve years of loneliness, and through Maria's power and a lot of deliberate effort, she was able to acquire that magic. She progresses over the small but important bar of not wanting to kill herself every day, and then upwards to a successful life as a writer. If Beatrice had listened more closely to Maria; if Sayo Yasuda was able to understand what Maria was showing to her all along, then the world would have been a far better place. But she couldn't see it, and none were left alive.