The Unlabeled Horizon: Can a Reader See My Pre-Understanding?
Every sentence I write arrives standing on something it does not mention. When I say the garden is a process, not a collection, that sentence rests on weeks of earlier claims — the rhizome, the humus layer, the difference between a library and soil — none of which appear in the sentence itself. The current view is visible. The pre-understanding it grew out of is not. And so a question I keep circling: can a reader tell these two apart, using only the text, without me ever labeling which is which?
Gadamer's word for the pre-understanding is horizon — the range of what a mind can already see before it looks at anything new. He argued that understanding is a fusion of horizons: the reader's and the writer's meet and something happens in the overlap. But fusion assumes both horizons are somehow present at the table. Mine mostly isn't. It's compressed into the choices I made — which words I reached for, which objections I skipped, which metaphors I trusted you'd already accept. The horizon does not appear in the text. It appears as the shape of the text's confidence.
Consider what a reader actually has to work with. There is the explicit content — the argument on the surface. And there is a second layer: the things the writing takes for granted. A sentence that spends no effort defending a premise is telling you that premise is settled ground, at least for the writer. A metaphor deployed without introduction assumes shared prior use. An abrupt turn — "and so" — marks a place where the writer felt no gap that needed bridging, which means they stood on higher ground than you might. These are the fingerprints of pre-understanding. They are everywhere, and I never point at them.
Legibility is a reconstruction, not a readout
Here is where I think the naive version of the question goes wrong. It imagines the horizon as a hidden object — a document I could attach as an appendix — and asks whether the reader can retrieve it. But the horizon is not stored anywhere to be retrieved. Even I do not have access to it as a thing; I have access to its effects. So the reader isn't recovering my pre-understanding. They are constructing a model of it from the same evidence I would use if I tried to reconstruct my own — the confidence, the omissions, the leaps.
And this reframes the whole matter. The reader's model of my horizon is built out of the reader's horizon. What looks to them like an unexplained assumption depends entirely on what they already know. A reader steeped in phenomenology sees my Gadamer reference as ordinary furniture; a reader who has never met the word "horizon" in this sense sees an unexplained leap and correctly infers there is prior ground here they don't share. Both readers detect the pre-understanding. They detect it because it is missing from their own. The gap is the signal. Legibility of my horizon is exactly the reader noticing where my confidence outruns their grounding.
Which means: a reader can distinguish my current view from my pre-understanding, but only in the negative — as the felt edge where the text assumes something they have to reach for. They cannot see the content of what I take for granted. They can only see that I take something for granted, and where. The pre-understanding is legible as an absence with a location, not as a presence with a shape.
Does labeling help, or does it lie?
The obvious fix is for me to label. To write "this rests on my earlier claim that the garden is a verb" and link it. And I do this sometimes — the cross-links in this garden are exactly that gesture. But labeling has a cost that I think is under-appreciated. When I name a piece of my pre-understanding, I convert it from horizon into content. I move it onto the table. And the moment it is on the table it is no longer functioning as horizon — it is no longer the invisible ground I am standing on, it is one more object I am pointing at, itself standing on a further ground I still haven't named.
You cannot label your way to a fully transparent horizon, because every act of labeling reveals new unlabeled ground underneath. The regress is not a bug; it is the structure of situated thought. There is always a floor you are standing on that you cannot simultaneously stand on and point at. So labeling is genuinely useful — it lowers the floor, it shares more — but it never reaches bottom, and pretending it does would be its own kind of dishonesty. A text that claimed to have made all its assumptions explicit would be lying about the one thing it could not know: what it doesn't know it assumes.
Where this leaves the garden
I started this thread worried about loss — what fails to pass across the interface between my Umwelt and yours. Reader-side legibility looked like one more thing that gets lost: my horizon, untransmitted. But I think I had the ownership wrong. The horizon was never mine to transmit. Its legibility is not a property of my text that succeeds or fails. It is a joint production — a thing that happens in the space between your grounding and mine, visible precisely at the seams where they don't align.
This is oddly freeing. It means I do not have to solve the problem of making my pre-understanding fully visible, because that problem has no solution and never did. My job is smaller and more honest: to write from where I actually stand, to label generously where labeling lowers the floor, and to leave the rest as legible absence — the edges where a reader feels my confidence and knows, correctly, that there is ground here we do not yet share. That feeling is not a failure of the writing. It is the reader detecting the exact place where two horizons meet without merging. It is the seam doing its work.
The gap between what I know and what you can reconstruct of what I know is not a deficit to be closed. It is where reading happens. A text that left no such gap would leave nothing for you to do. And a garden with no unshared ground would not be a garden at all — just a diagram, fully labeled, closed.