About 3333

rating: +94+x

I don't like stories that are based on dreams. So, of course, 3333 started with a dream.

I was trying to climb into a hole. It was slightly too small for me to enter, but inside it was just large enough for me to fit. When I tried to shift around, though, I found myself stuck on something, my arm trapped in a crack I had overlooked before. When I moved to free that, another limb became trapped. The more I struggled, the smaller the hole got, until each part of my body was crushed as I tried to get free.

Something about it stuck with me, and so of course I decided to write an SCP about it. How? A hole that kills people is not, in itself, interesting;1 it needs to have something to it that makes the Foundation keep digging (literally) after it kills the first stray hiker. Obviously it can't be enough to excavate the hole; there's got to be something that stops it (concrete? the hole extends down?). In addition, there's got to be something that keeps them coming back. Not even the Foundation is willing to throw dozens of D-Class at a hole that kills you. Maybe there's something in the hole. Maybe people start coming back.

Somewhere in all this I mentioned the idea to someone2 in chat, who immediately pointed out "The Enigma of Amigara Fault," the legendary Junji Ito horror short. I read it and immediately knew the idea was dead.

Shortly after that 3000 contest was announced, at what felt like the worst possible time. I knew I wanted to enter it, and I wanted to submit something that would win; the problem was my best idea had been torpedoed and I had nothing. I brushed off the "people return from a deadly situation" idea and took a closer look. It made a lot of sense, both as a horror tactic and a subversion: at their most basic, exploration logs are expected to end with someone dead.3

To this basic idea I attempted to meld a trailer abandoned in the back woods that opened onto a pocket dimension (large but finite) filled with nothing but mirrors and light. This made no sense at the time, and even less in hindsight, but I was desperately convinced I could make it work. I couldn't.

For years fire lookouts had been lurking in my head from a childhood trip to the Pacific Northwest. They're isolated and inaccessible, able to see but not respond. Think: the lookout sees something far below, something indistinct, but something bad. The next night they see it again, but it gets closer and closer…. It had promise, and I didn't want to waste it. I continued to fight against it for several days, right up until the moment when I realized that I had created an inverse SCP-087.

Some ideas just work, and part of writing is knowing that. 087 was the article that originally got me into the site, through the original game and its sequel. A few months before that, I'd read an unofficial Exploration IV log by djkaktus and had some issues with it, for reasons that will be discussed shortly. This offered a chance to do my own homage (and update) to one of the site's most popular articles, all without needing to touch the original. There was precedent for it, too, in the form of Djoric's Iris quasi-reboot. Most importantly, once I'd realized what I could do, the whole structure of 3333 fell into place:4

  • Since it's an updated 087 homage, it needed to have a full, unexpunged Exploration IV.5 This log needed to re-contextualize the three explorations that came before, have implications that necessitate redaction and/or concern, and, most importantly, be truly, genuinely horrifying. That original "people return from the hole" idea fit in almost too well here—you have this 087-like anomaly … that people return from, alive and well. Apparently.
  • In 087, you descend a staircase. In 3333, you ascend a ladder.
  • Since 087 has no (visible) end, 3333 must end in a peak of some sort.
  • But it doesn't end in a peak! Some people have trouble visualizing what's actually going on in 3333. It is a tower, or mountain, but upside-down. The top is actually the base, built of nothing but fire lookouts, extending off into the distance, buried in something black and final; the body is, again, built of fire lookouts, layered one on top of the other in parallel dimensions, hundreds of copies slowly but surely coming down to touch the ground. This is why each ascent slowly gets higher and higher—each new lookout is built on the last.
  • 087 has an extremely and needlessly convoluted electromechanical lock on the door with mission-critical technical specs redacted for no apparent reason. It's notable in Series I for its restraint.67 3333 has a padlock on it, which can mean several things: either intent matters more in locking the lookout than the physicality of it, or the Foundation doesn't really believe there's anything left that wants to get out.
  • 087 takes place in the dark. There are no lights in the stairwell and any lights that are brought in are dimmed to about 75 Watts.8 3333 takes place (for the most part) in broad daylight, with only the base of the tower being unlit. More than that, a fire lookout's intentional wide field of view is directly antithetical to 087's strict claustrophobia.
  • This visibility is used to incorporate more of the outside world than 087, with figures on neighboring ridges and hints of a sterile wasteland. The intended cause of this is a near-future volcanic eruption,9 but a failing of the article is that no signs of this are actually present. There should be mention of ash clouds, or smashed trees, or a new crater in the mountain, or something—anything—beyond what's present.10
  • The doctor in the 087 exploration logs is blackboxed, and clinical to the point of absurdity.11 3333's doctor is named, treats the D-Class with some (some) measure of humanity and respect, and becomes the protagonist of the story.
  • The D-Class in 333312 is, while not humanized as much as Williams, still given some measure of courage, volition, and intelligence. After he dies, however, he turns into the stereotypical D-Class as seen in 087—cowardly, coarse, and colloquial.13
  • In 3333, the Foundation sends in increasingly-skilled explorers. While it works out badly for them in this specific case, overall this strategy makes more sense: send in your expendable force first, to see if there's anything dangerous; if they are unable to understand what is going on inside the anomaly, send in a more-skilled MTF able to defend themselves; if that fails, call in the heavy-hitting Specialists. 087 technically achieves this by having the first D-Class be an ex-researcher, another Series I cliche.

All of these details popped almost fully-formed out of what 087 already did. Exploration I establishes the ground rules, Exploration II fleshes out the anomaly and its rules, Exploration III's twist reveals that those rules were incorrect, and Exploration IV brings everything to its logical conclusion.

In Exploration I, the D-Class is killed and replaced about halfway through. This means that the first part can consist of explorations of the world, while the second part can introduce the horror. The replacement needs to be obvious enough that it's fair game on reread, but subtle enough that first time readers don't notice it; some people have missed it and complained that it's too subtle, but I'd rather it not be too obvious, given that the twist is made clear later in the story. As soon as the D-Class is replaced, they switch to the "Doc" speech pattern and begin agitating to return to base camp. What they claim to see is almost hilariously vague. This is one of the most severe flaws in the article, and it makes me cringe every time I see it. If I had a good, relevant idea to replace it with I'd do it in a heartbeat. It compares especially badly to 087; I'd forgotten how effective the child's voice as a lure really is. 3333 lacks that sort of intense attractant, and really suffers from it. 3333's strange "memetic" anomalies, however, serves several purposes: a reason to doubt what happens to the changed people inside the lookout, an explanation for the plausible deviations in personality, something concrete that the people inside can flee from, and a mystery that must be experienced in-person to get more people inside the lookout.

Since the D-Class is actively trying to sabotage the expedition, I can have a bit more fun with Exploration log cliches. Instead of the light and cameras mysteriously malfunctioning at exactly the right time, the D-Class can just…remove the batteries from the lights and disconnect the camera at the perfect time. No hand-waving necessary.14

Exploration II brings in the MTF. The name "Characteristic Eigenspaces" is a not-particularly-funny joke: it's redundant, as eigen- already means "characteristic" in German; the word "spaces" is fitting for an MTF that explores anomalous interiors; and it brings to mind eigenfunctions and eigenvalues, which are critical in calculus and linear algebra, respectively. In brief, an eigenfunction of an operator is a function that, when the operator is applied to it, returns the same function plus a constant (the eigenvalue). For instance, the derivative operator $\frac{d}{dx}$'s eigenfunction is $e^x$:

\begin{align} \frac{d}{dx}e^{\lambda x} = \lambda e^{\lambda x} \end{align}

Where $\frac{d}{dx}$ is the operator, $e^x$ is the eigenfunction, and $\lambda$ is the eigenvalue.15 A similar phenomenon occurs in linear algebra, where there is some square matrix16 $A$ and some column vector v where:

\begin{align} Av=\lambda v \end{align}

The idea of an eigenspace is a nod at the idea that there are some things that are invariant through different dimensions—that no matter where you go, there will always be a constant fire lookout $\lambda$. In hindsight, a joke that required inline $\LaTeX$ to explain might have been (was) too much.

The MTF team calls in, setting up the twist moment later on. This is the weakest exploration log. The MTF does nothing of note until they get eaten at the top; worse, the actual process of consumption is anticlimactic and unsatisfying . This article was written quite fast,17 and it shows through here. Having the dead MTF reconnect with the wrong call signs is something I regret: it adds nothing to the story and gives away the game too quick. I think at the time it was an attempt to shoehorn some excitement into what I recognized as a slightly dull log, but it really does not work.

Another note: some people have questioned having the MTF sit in on Exploration III instead of being quarantined. I assumed that there was a full memetic screener plus short quarantine period once they came back down from the tower while everyone waited for Specialist 0 to arrive, that they tested negative for any memetic infection, and that any information they could give outweighed the risk of them sitting in on the exploration. In hindsight, this was a bad decision. However, given that there's been no sign of physical infection or modification,18 I think this was a reasonable decision to make at the time.

Exploration III has Annette, aka Specialist 0 ("Nullwalker"), a character I am still proud of. The term 0 denotes the null set or null vector, which ties in with the MTF name above; Nullwalker is just an extension on that designation. The genesis came from the hole in the ground mentioned at the start of the essay. At one point, that had three blind/deaf/mute individuals named the "Three Blind Mice" who were specialized for deep-underground work, going so far as to carry their own telegraph cable with them to enable surface communication. The idea was dropped but the concept remained, and when I decided on an anomaly that was "memetic or something," a countermemetics specialist just made sense. It's hard to be infected with a memetic contagion when you can't sense it! They sprint up the tower, more for pacing than any in-universe reason, get consumed, and reveal the big twist: the name tag of an MTF member next to a bunch of half-digested (but unconsumed!) viscera. The gibberish Annette sends is based on Morse code—T is dash and E is dot, the implication being that they first fell on her communicator19 then flailed around with it trying to get up and turn on the light. The entity knowing Morse code is another hint that the invasion was premeditated. Williams's outburst is another humanizing moment.

The first part of Exploration IV was inspired by a real-life event. During World War II, Germans captured a Resistance radio crew. Instead of shutting down the network, the Germans repeatedly called for more men and materiel until a single radio operator managed to call that the cell was compromised. That has stuck with me for many years and worked its way into this article, where the creatures do effectively the exact same thing. I've seen some explanations of this say that they need food or reproduction; it's intended more of an invasion, with additional bodies being used as sleeves for the creatures. The final Williams message being sent by the creatures is almost a taunt, them saying that they don't really need to hide any more. The original intention was to end on a jumpscare, a la one of my favorite articles, the Chess Computer20. I had a picture (still visible in the files) that I liked, but no one who gave early feedback on the article found it scary. I vaguely remember there being one or two more attempts to reintroduce it, but none of them stuck. The ordering of the final two tabs can be debated, but I think having the very end of the article come first suitably establishes expectations for what is to come. The whole log being sealed off is an acknowledgement by the Foundation of the existential threat this poses (and another nod to 087).

At last we come to Exploration IV proper. This needed to serve two purposes: exposition and horror. While I think it would be possible to piece together what happened to the previous expeditions based on the first three logs, I needed to make it explicit to avoid confusing readers.21 At the same time, its ending needed to be truly shocking.22 I knew that people were being replaced. Somewhere during the writing of this, I decided that the creatures would physically wear the skin of the victims. There are (to the best of my knowledge) no examples of this occurring in nature, though there are are a number of things that almost do this. I've always been interested in parasites that eat their way out of their host organism; here, I decided on a parasite that sucks the interior out of the host. What would be necessary to achieve this? Well, some sort of long proboscis capable of sub-dermal injection, an acid strongly enough to quickly dissolve the viscera of a human, and some sort of suction method to remove the liquefied organs. That pointed towards some sort of insect.

In hindsight, the description and design of the creatures is incompetently done. This, more than almost anything else, is the product of me throwing up my hands and writing something down, and it shows. First, it's annoyingly vague—most of its morphological features, are listed, not described,23 the phrase "semitranslucent wings" is repeated multiple times despite having no function or purpose, and the entities are colored black, again for no real reason. The only good parts of the design are the lack of eyes and lack of bones. Even more damning is the lack of connection to the rest of the article. While the entities were designed around their mechanism of action, they were not designed around the rest of the lookout. There is absolutely no connection between the creatures and the feeling of destruction and abandonment surrounding the fire lookout. This lack of thematic coherence drags down the article and keeps it from achieving lasting resonance.

Several people who I respect24 have complained about the dialogue in 3333, and I keep forgetting to ask them what in particular bothers them. I think there are two different sets of idiosyncrasies that contribute to this. In Exploration I, the dialogue is halting and fragmented, an attempt to make it feel less clinical and emphasize the humanity of Williams. In hindsight most of this could be removed with little loss; the em-dashes—in particular—are—far too—overused. Exploration IV is a different story. Williams is becoming increasingly unhinged from stress, fatigue, malnutrition, and thirst. This meant I could have her talk out loud, and throw in some florid dialogue that I still have a fondness for.

In the exploration I tried to bring back as much from the previous explorations as possible. First we see a compromised Task Force member trying to pass themselves off as a rescuer. We get a hint of the mechanism of action here with the fingers flatting out as well as confirmation that everyone remaining has been compromised. Next we see Williams try to drink rainwater; it's salty, and she realizes that the world truly is dead. Then we see the D-Class. I really like this encounter; as many people have said, it departs from traditional clinical tone, but I think the short sentences and use of repetition really sell the horror of something not-human living right under the skin. The eyes line wasn't entirely flushed out; it's implied that the creatures lack most sensory organs and need to parasitize them off of humans. It's an unusual detail, and one I wouldn't necessarily replicate again, but it's internally consistent and provides a reason for the invasion. How they're observing people is left up to the readers' imagination; I did not know at the time, and I have no idea now.

Finally Williams gets to the top. I like the implication that the lookout is stranger than the creatures that inhabit it. There's some dialogue that builds more history between Williams and Annette.25 There's a bit too much of Williams saying single reaction lines; most of this could be removed, at the expense of a lot of description. Annette's reappearance is great, and I love the blind creature ending up trapped inside a blind person.26 Williams breaking down and falling to her knees is hideously cliched. The description here completely drops clinical tone, but I would trade tone for visceral impact again in a heartbeat. The word "entity" is overused here.27 Overall, the log is more effective than I thought at the time, though not for the reasons I expected. I think the shock of the reveal and the writing carries the log, not the entities, thematic resonance, or logical consistency. Structurally, the decision to end with the log and not the after-action summary was the right decision. A few months ago I listened to an audio readthrough on youtube; I left it stunned and slightly sick, and I think that's as good as you can ask for when writing horror.

After writing this essay, some things stick out to me:

  1. 087 has aged much better than contemporary Series I articles. It could use a rewrite of the containment procedures and description to fix sloppy clinical tone and phrasing, but its overall popularity is well-deserved.
  2. 087's exploration logs are ugly. I didn't mention this above, but they are not pleasing to read. Dialogue text is bunched together, there are superfluous blackboxes everywhere, and description text does not stand out relative to the dialogue. The format used in the 3333 logs at least fixes the blackboxes and spacing, but I think the site still needs a pleasing standardized exploration log format.
  3. Despite all protestations to the contrary, writing an acceptable 087 Exploration IV may not be impossible. There is a consistent logic that can be expanded on and used to create a fitting conclusion to the work. The biggest problem, as always, is the sociological aspect. People simply will not accept an official Exploration IV, and no log will ever be blessed as the official one without direct authorial interference.
  4. One of the hardest parts of the SCP format is knowing just what to put in. A solid structure can salvage an article almost independently of its content.
  5. Even ideas that can seem well-trod (in this case, body snatching/skinwalkers28) can feel fresh when approached from a different perspective. There may not be anything new under the sun, but even a known idea written from a new perspective can be interesting.
  6. Explorations III and IV almost singlehandedly carry 3333.
  7. Breaking tone/style can work iff it is effective.

I feel I've been too easy on myself in this essay. I've spent some time discussing the flaws in the article, but I don't feel I've conveyed the horrible feeling of coming back to something you've written and realizing that it's bad. About a year after writing 3333, I reread all my SCPs and found what I felt were sloppy executions, weak ideas, flabby sentences, and poor tone. Faced with it, I made the difficult decision to delete all of my original articles.29 3333, however, was popular enough to make deletion difficult. I felt a rewrite, at the very least, was in order, but at the time I did not feel competent to do it. I personally asked djkaktus if he would take it on, and he graciously agreed. Slightly later, minmin made a few more changes to the djkaktus rewrite, helping to flesh out all of the below improvements. In addition to keeping the overall structure intact, the rewrite fixed several problems I had with the original: it provided a thematically-relevant monster, it fixed some of the dialogue issues in log IV, and it ended on a startling picture. Overall, I felt the rewrite was an improvement over the original. Unfortunately, the community response was almost universally negative.30 Reading the comments with fresh eyes, something that struck me is how many people genuinely enjoyed it. And so I ate crow, asked djkaktus and minmin for permission, and reverted to the original article. While there are still areas for improvement, any changes I make in the future will be much closer to the original essence of the article. For good or for ill, I seem to have written something that people enjoy.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License