·4 min read
The narrator who tells you he was wrong before
A Hudson Valley experience built entirely out of found documents, where the person handing them to you opens by admitting his last lead was rubbish. Notes on why that admission is what makes the rest land.
An experience in the Hudson Valley begins with a note from a man called Bishop. Before he tells you anything, he apologizes.
He says the last sequence he sent you, the one with the giant carp, was bad information. He was having cluster headaches that week and was not doing his diligence. This one, he says, is the real deal.
That apology is the most load-bearing sentence in the whole piece, and it took me a while to work out why.
Found documents do something first-person cannot
Nothing in this experience is addressed to you as a character. There is no narrator explaining where you are or what you should feel. What you get instead is a stack of things other people made for other reasons.
A letter from Bishop, an amateur, chasing a figure he is not certain exists. A 1941 note from the superintendent of a state hospital, preserving a run of patient cases he cannot explain and does not want to be laughed at for. The recordings themselves, five patients describing a late-night visitor in their own words, decades apart, each one certain and slightly embarrassed. And an engineer's introduction explaining how the transmissions are being kept running at all.
You are not being told a story. You are reading someone else's correspondence and drawing conclusions they did not draw.
This is a very old trick, and it works outdoors for the same reason it works in a novel: it puts the reader in the position of investigator rather than audience. But there is a specific reason it suits walking. A document does not have pacing. You can stand in the wrong place, get distracted, come back tomorrow, and a 1941 clinical log is exactly as convincing as it was. It is not performing. It is just sitting there being a document.
Why the apology is the whole thing
Here is the problem with any story delivered as if it were real. The player knows it is not. You cannot out-argue that, and trying makes it worse. The harder a piece insists on its own truth, the more it feels like a piece insisting.
Bishop solves this by being obviously fallible. He got the last one wrong. His wife thinks he is being led on a multi-decade goose chase by a troll with a sick sense of humor. He appends his theories to things and admits he cannot make sense of the ending. At one point he explains that he found a crucial clue because he needed to take a leak.
None of that undermines the fiction. All of it holds the fiction up.
Because the moment a narrator admits he is sometimes wrong, the things he is sure about start carrying weight. You are no longer being asked to believe a story. You are being asked to evaluate a source, which is a much more interesting thing to be asked, and one you can genuinely disagree with. Bishop even tells you he thinks your speculations are often wrong. That single line does more for his credibility than a hundred words of atmosphere.
The clinical voice does the rest
The hospital records are written flat. Diagnosis, event date, symbol found, outcome. Nothing in them is spooky. The superintendent is careful to say only what he observed, in the register of a man who expects to be read by colleagues who will judge him.
The patients' own accounts are where the warmth is, and they are warm because none of them are frightened. A woman describes a visitor she talked to about how the building groans when the steam comes on. A telegraph operator watches a stranger tap a rhythm on the bed rail and recognises the cadence. A quartermaster lets a man help him tidy his locker and learns they were at Verdun months apart.
Five people, decades apart, each visited once, each calmer afterwards. Nobody in the documents ever says the word that would ruin it.
The ask at the end
The last thing this experience does is ask you to do something small and physical with a piece of chalk. Not solve a puzzle. Not answer a question. Contribute.
I will not say what, because it is the point of the whole walk. But the reason it works is everything above. By the time you get there you have read five people's accounts, formed a private opinion about whether Bishop is credible, and decided for yourself how much of it you believe. Being asked to add something to that is very different from being asked to complete it.
What I would tell someone starting one of these
Write your narrator as a person who could be wrong, and then let them be wrong about something, out loud, early.
It costs you one paragraph. It buys you every sentence after it.