The Conditions of Coexistence

One invitation, one shared afternoon, no assigned activity beyond bringing whatever helped a person slow down. What I watched happen split cleanly in two. Some people read, painted, or sat in silence, alone or near others without speaking. Most did the opposite — they danced, met strangers, exchanged numbers, moved from one conversation into the next. Both versions ran the entire afternoon, in the same space, at the same time, without either one winning.

That shouldn't have been the likely outcome. A setting this loosely defined is exactly the kind that research on shared environments would expect to collapse into one dominant use, the way behavior in a physical space is usually shaped more by the setting itself than by whoever happens to be standing in it. This one didn't collapse into anything. It held two, at once, for hours.

At first, comparing two experiences seemed like the easy part.

The behavior was already visible — who danced, who read, who stayed quiet. Comparing that looked close enough to comparing experience to keep going.

It wasn't close enough.

The question kept resisting a clean answer the further I pushed on it. Were the readers and the dancers having different behaviors, or different experiences? If experience, different in what sense — different in what they noticed? Different in what they wanted going in? Different in how it felt while it was happening? Different in what they'd remember a week later?

Each answer pointed somewhere else. Behavior was only ever the visible trace of something else, not the thing itself. Feeling wasn't accessible from the outside for anyone but me. Expectation belonged to before the gathering, not during it. Memory belonged to after — and memory, it turns out, isn't even a recording of what happened. It's a reconstruction, built from a couple of standout moments, not the whole stretch of time.

None of those were wrong, exactly. None of them were the whole answer either.

Before the investigation could go any further, it needed to answer a more basic question first.

What do we actually mean by experience?

The answer that held up treats it as none of those things on their own. Experience isn't what an environment gives a person, and it isn't what a person brings to an environment. It's what happens between the two — made through the encounter itself, not delivered by one side to the other — and it keeps happening before an encounter, during it, and after. Adopted here because it's the most useful way to study what follows, not because it settles a debate that's still genuinely open elsewhere.

Behavior stays in the picture from here on as the visible trace of that relationship, not as a stand-in for the whole of it.

Several lines of thinking get partway there.

One says a shared space behaves less like a blank slate than like a script — that people conform to what a setting seems to call for, more than personality alone would predict. If that were the whole story, a setting this unscripted should have let one default mode take over. It didn't take over anything. It held two modes for an entire afternoon.

Another says people bring their own needs into any shared setting, and what they do with it depends on which need is active — some are looking to restore themselves, others to connect with someone. That fits the split almost exactly: the readers and painters read as people meeting a need for rest; the dancers and the conversationalists read as people meeting a need for connection. It's a strong account of why each group chose what it chose. It has nothing to say about why the setting let both needs get met side by side instead of one crowding out the other.

A third says the same physical space simply offers different people different things to do with it — a blanket can be a place to read or a place to sit near a conversation, depending on what someone is already looking for. That explains the selection. It doesn't explain the coexistence either.

The sharpest gap shows up in a fourth line of thinking, about what happens when a group gathers around a shared rhythm — music, motion, mutual attention. Groups like that are usually expected to generate enough energy to pull bystanders in, the way a dance floor tends to grow rather than stay contained. That's exactly what didn't happen here. The readers stayed readers. Nothing pulled them in, and nothing pushed them out.

Between all four, most of the divergence is well accounted for. None of them explain why it held.

A City Sidewalk

The writer and urbanist Jane Jacobs spent years watching city sidewalks do something similar, at a much larger scale, and documented both when it worked and when it didn't.

Her successful case: dense, mixed-use city blocks, home to people heading to work, to a shop, to their own front stoop — all moving through the same sidewalk for reasons that had nothing to do with each other. Nobody was there because of anybody else. Cycling through together anyway, they produced what she called a sidewalk ballet — a shared, self-sustaining order with no single choreographer.

Her failed case sat a few blocks away, in postwar public housing towers surrounded by open plazas anyone was technically free to use however they liked. The plazas stayed empty. Without people passing through for their own unrelated reasons, there was nothing to sustain shared use — and empty didn't resolve into calm. It became unsafe.

Permission was equally present in both cases. What was missing from the failure, and present in the success, was enough people arriving for genuinely independent reasons to keep the space populated by more than one purpose at once.

An Open-Plan Office

A more recent, better-measured case comes from ordinary office buildings. When companies remove the walls between desks, hoping to blend focused work and spontaneous collaboration in one space, researchers tracking real employees found the opposite of what was intended: face-to-face interaction dropped by roughly two-thirds, while messaging and email rose to fill the gap. With no way to actually separate the two modes, people didn't blend them. They withdrew from both, and improvised private signals — headphones standing in for a closed door — to claim whatever boundary the room itself refused to provide.

This is the case that looks the most like the gathering and behaves the most differently from it. Both are open, unzoned, self-signaled spaces meant to hold more than one mode at once. But a phone ringing at an empty desk reaches every desk regardless of who's wearing headphones, the way a dancer's noise never had to reach a reader with enough lawn between them. Divergence had nowhere to go that didn't cost the other mode something. It broke.

A Public Library

A public library is the sharpest test, because on paper it should be the least likely of the three to work. It ranks its own norms openly — quiet is the stated default, not one option among equals — and still holds silent study, loud children's programming, computer use, and casual browsing in the same building, at the same time, without collapse.

What makes it work isn't an absence of hierarchy. It's zoning — distinct rooms, different acoustic treatment, different furniture, often a person on staff actively managing the boundary between them. A library doesn't dissolve the difference between quiet and loud. It gives each one somewhere to happen without reaching the other.

Two guesses came first, and both had to be abandoned once they met more than one case.

The first was permission — that a loosely defined setting is what makes room for divergence in the first place. It didn't survive the library, the least loosely defined of the three environments and the one that worked best. Permission turned out to be necessary only in the trivial sense that nothing coexists in a space that forbids it outright, and it explained none of the actual difference between the cases that worked and the one that didn't.

The second guess, drawn from the office, was that no mode could be allowed to outrank another. The library broke that one too. It ranks quiet above loud, openly, and still works.

What survived, across a city sidewalk, an office floor, and a library — revised twice before it held — is narrower than either guess. Coexistence turned out to need two things, not one: people who came for genuinely independent reasons, and enough room for each version of the experience to unfold without costing the other anything. Every case that worked had both. Every case that failed was missing at least one.

• Whether the two conditions are actually independent of each other, or whether one only works because the other is already in place — every case here had both present or both absent, and none of them separate the two.
• What produces the shared, relaxed undertone that seemed to run underneath both modes at the original gathering, even though nothing here tested for it directly.
• Where a person's need for rest or for connection actually comes from, before they ever arrive — every account here treats that difference as a given, not as something to explain.
• Whether any of this holds outside a physical room at all — inside a product, a conversation, or an interaction with a piece of technology built to hold more than one person's needs at once.

The first of those questions isn't just a loose thread. It's the difference between designing something that needs one condition and gets partial coexistence, and something that needs both together or gets none — a distinction that matters to anyone actually building a room, a product, or an experience meant to hold more than one person's version of it at once. No case gathered here can separate the two conditions from each other. Testing that directly, rather than finding it, is what Chyless World exists for.