Paired journaling experiment, seven-day results

Love that you collated at the end rather than as you went — that's the bit most of these experiments skip. The night 3 line is interesting to me because a flat horizontal that "won't tilt" is such a specific thing to write twice. But I'd want to know: did either of you talk about horizons or planes in the days before you started? Priming is the obvious culprit.

Fair, and I can't fully rule it out. We both read the same intro material months ago, so "plane" and "horizon" are baked into how we describe things. That alone probably nudges us toward the same vocabulary. Which is exactly why I only really trust the line because of the tilt detail — we both wrote that it stayed level when we expected it to move. That's a more specific note than just "horizon."

Two things. First, seven nights and two of you is a tiny sample — I know you said that, but I want it stated plainly so nobody skims this and repeats it as a result. Second, the "green" one I'd drop entirely. It's April, you're both surrounded by green, and green is one of the most commonly reported colours in any relaxed eyes-closed state. That leaves you basically one interesting data point, which is honestly a fine outcome for a week.

Something worth naming: when you know the goal is "find overlaps," your eye finds them. I do this myself — I'll read her "shelf" and my "horizon" and feel a little jolt, and the jolt feels like evidence. It isn't. It's confirmation bias arriving on schedule. Not saying nothing happened. Saying the feeling of significance is the least reliable instrument in the room.

This is such a lovely careful post though. Even if it's all coincidence, the discipline of writing before you speak to each other is worth it for its own sake. I might try it with my sister. If nothing overlaps, that's still seven diary entries I wouldn't otherwise have.

Worth flagging the terminology here, bc the Meridian papers are actually careful about this distinction and we usually aren't. They separate correspondence — two states that happen to share features — from contact, which implies a channel between them. What Bram has is, at most, correspondence: a couple of shared features across paired records. Nothing in the method could show contact, and I don't think Bram is claiming it does.

To be clear I'm citing the Institute as the source of the vocabulary, not because they endorse a word of this. They'd file the whole experiment under "anecdote" and they'd be right to. But the vocabulary is genuinely useful for keeping ourselves honest. If we all said "correspondence" instead of quietly meaning "contact," half the overclaiming on this board would stop.

Yes — "correspondence, at most" is exactly the ceiling I'd put on it. Editing that into how I talk about it going forward. Thank you.

Coming late but wanted to add a small design tweak for anyone repeating this: seed a few "decoy" nights where you deliberately don't sit, and still write a diary. If your overlap rate is the same on decoy nights as on practice nights, you've learned something honest — probably that two friends in the same season just write similar diaries. That's the control I'd add before I read anything into a horizon line.

Signing off this one having changed my mind slightly — not on the result (still basically noise) but on the value. I'd start the decoy-night version with someone next month. If nothing separates practice nights from decoy nights, that's the finding, and it's a good one. Thanks for doing it out loud, Bram.