It seems appropriate to begin these concluding remarks with the most secure and reliable data before a later transition outward.

The majority of the data in this collection is personal testimony and autoethnography, a research method in which a researcher tries to make sense of their own experiences through observation of themselves in a specific social context. This is as much the "roommates" writing about each other as me writing about them, making curatorial decisions about relevance and comprehension. The "Un-Monster" blog was public information with a negligible viewer response. Many of the text messages and notes in the "Brandon and Finn" database were written by one of the paired participants and intended directly for the other. "Sam" notes were supposedly not written with anyone else in mind, but their patterning and intimate knowledge of Tyler's blog content makes "Sam's" motivations ambiguous.

However, as the in-text notes suggest, the connections between the files seem deliberate. The information is often not supplemental but a complete omission. Important information appears in multiple places or nowhere at all. The redundancies throughout the text are so pervasive that their possible omission would result in a distortion of individual character. Without them, any practical community-centered utility based on these experiences becomes impossible.
If the purpose of this collaboration is some attempt at reality, then we must admit the limitations of that attempt. 

The text as a whole exhibits well-established defects. The participants use an egocentric viewpoint which lacks consideration of other perspectives. They discuss emotional fantasies with the same tone as material observations. They situate themselves as witnesses to (causes of) verified events with no possible human-centered cause. 

The files contain some information with an instructive tone, a tone that clashes against the personal and self-referential subject matter.

The primary cause of the roommates' repeated failures is a lack of empathetic imagination. They had multiple opportunities, over a minimum timescale of thirty days, to maintain a positive relationship with each other and identify the source of their "effects". They refused many of these opportunities until the relationship as a whole was no longer viable without extreme intervention. 

The roommates often face trivial, distracting versions of their larger personal struggles. Their lack of skill when solving the miniatures means they are ill-prepared at solving their later replicas of increasing scale.

These insights are only possible with the addition of context and emotional distance from the conflicts as they happened. Like other contemporary disasters of the late twentieth and early twenty-first centuries, personal accounts are mediated through social interactions and technological data channels. This is a concern that informed our entire approach to the project. We received multiple testimonies which we have reviewed and discredited outside of what is shown in the collection. It is not possible to wholeheartedly trust events that a single participant recorded due to their personal biases. But these biases are also susceptible to interpersonal influence. Even when events are corroborated, the observing participants (especially in the main collection) resort to figurative language. E.g. One potential outside source for the project based their testimony on a live news report they remember overhearing while preparing a meal for themselves and their spouse. These are well-known sources of cognitive noise: a news report which condenses and frames possibly accurate information to fit within a specific format and tone, itself largely based on personal testimony; political biases of eyewitnesses or other commentators in a news context; viewership experiences which inform biases towards or against news media; cultural memory which repeats over time across contexts and distorts into simulacra; classic audio-based noise from movement and activity in a given space; affect changes based on interaction with a loved one; social rituals based on meal preparation and news consumption itself. Another potential source at first claimed eyewitness testimony and included details that were only available in subsequent safety bulletins. Such concerns are supplemental to the already-documented noise present in the primary collection, the events of which are now over two years passed.   

At the same time, these sources of noise are factors during the creation of memories. Even sources who had "eyewitness testimony" and were immediately present for events are often distracted and have too wide a variety of experiences to make any meaningful conclusions. For example, see [Oliver's] first series of 10 community interviews.

The abundance of primary source material and the comparative lack of public secondary sources make verification difficult. Participants do not experience everything, remember less, write down less than that, and keep less than that after further consideration. Data lost in this decay is irretrievable. If verification is only possible through corroboration against other trusted sources, then the quality of truthfulness itself is inductive at its own peril. We can only verify what matches. Is the assumption that the participants are "telling the truth" in a consistent and relevant way based on a coincidence that their telling occasionally matches reality?

The files available to the public are likewise replicas of real objects. In the "reality" of the events, none of the file collections or individual contributors has unequal weight. But all matters of library science as well as history, archaeology, anthropology, journalism, and the like which seek a definitive "what happened" seek a goal firmly rooted in ontology with useless epistemic tools. Much like a bird building a nest out of dirt or water instead of a more appropriate material, we have the experience, not as it was experienced but as it was written. In only this, we are successful.

This is a well-established constraint in the field. Oliver himself identified it early on during his work. He described the objective as "impossible and useless".

His work continued because this was not Oliver's only objective.

Oliver, you got emails from people who don't live here, who aren't even in the country, much less the state. They say they want to help you. Do you think they have any context for this disaster?

Oliver, you don’t consider how people think, or how they feel. That's what keeps you surrounded by unchanging words. Why bother with locks and gates and tangling roots? You know why you're here. You think you make a choice to stay here. Of course, you do.

 

Our success, while definitive, has coexisting constraints. The descriptions found in the “Wildland Animals” text suggest that Oliver’s metaphysical development is only possible once. His candidacy was dependent on certain social factors, which we will omit for his safety and well-being. His file processing tends to reduce subsequent exposure and hazardous affects by non-related parties to a negligible, non-zero level.  

The exact conditions are unknown but we have demonstrated their replicability.

Despite our exhaustive documentation, Oliver's exact methods for rendering text legible remain opaque. Our community will also continue to refine our crude model of Oliver's "burrow" procedure. His participation is not optional. Community members are currently being briefed on the topic. 
We speculate that Oliver's lack of personal attachments and demeanor as an interpersonal non-threat only further confirmed his eligibility. The contrast between "the roommates' " tense environment with no clear goal and Oliver's goal-oriented environment was one of many factors we took into consideration. 

There is no evidence to suggest the "disaster" has biochemical origins. Neither interpersonal exposure to the "roommates" nor exposure to the Wildland Animals zine are necessary prerequisites for the observed instability. We still encourage the use of extreme caution when interacting with the participants. Across [redacted number] other case files with potential exposure, none have shown evidence of "transmission", except for the "Oliver Whitney" case. Although we try to maintain a neutral tone and approach, our community lacks the necessary resources for the Roommates' long-term care. Our inability to identify the relevant transmission factors led to models based on social contagion. Further analysis suggests that this initial assessment is inaccurate. In comparison to the reliable analog replicability of the procedure, the discrete digital sections are not relevant to you.

Once "Oliver" demonstrated the ability to not only interact with artifacts but the "roommates" themselves, the process concluded. "Oliver's" presence itself is an effective counterstrategy to both exposure and development. There is no evidence to suggest the strategy has purely biochemical or sociopsychological origins. Again, this may be "something else entirely".

The library campus has become inhospitable during Oliver's extended residence. There are no current plans for reopening public access. 

In conclusion, we have unfortunately arrived back at where we started. We have looked for the truth. We have found nothing.