The word for it now is polycrisis: crises that stop arriving one at a time and start feeding each other, with no clear end in sight.

What that does to a person is subtler than the crises themselves. It turns uncertainty from a shortage of information into a property of the world. Complex systems are open, dynamic and self-organising, so the future is not merely unknown — it is undetermined. No quantity of additional data closes that gap, because the gap was never epistemic.

Most collapse talk is still about a wall: when we hit it, how hard, who is on the far side. This project is about the duller and harder question of what someone builds who has already accepted the wall and still has to be somewhere on Monday morning.

Three things it holds together

What the climate sciences can now say about weather nobody has records for. What anthropology says about presence — the capacity to remain a person while the order around you stops holding. And what a designed place can actually do with a body standing in it.

The answer this keeps arriving at is small and physical: build gardens. Not as consolation for the argument, but as the conclusion of it.

Small planted places, treated as urban equipment — at the same level of seriousness as drainage, shade and shelter.

Not amenity. Not decoration. Not a wellbeing feature bolted on at the end of a development. Equipment: the thing a city installs because people cannot do without it, maintained on that basis, budgeted on that basis, defended on that basis when the cuts arrive.

A garden of solace is designed backwards from what a body needs when the news is bad and the air is thirty-eight degrees. Shade that works at the hour it is needed, not at noon in a rendering. Water you can hear. Ground uneven enough to require something of the feet. Things at hand height that reward being touched. Sightlines that give both refuge and prospect. And enough enclosure that a person can say a true sentence out loud without performing it.

Why the word matters

Because "amenity" decides the budget line, the maintenance contract, and whether the thing survives the next administration. Heat, grief and dread are load conditions. We design for the other load conditions without apologising for it.

What it is not

Not a carbon offset. Not a substitute for care that should be funded properly. Not a claim that planting fixes anything upstream of itself. The ambition is local and should be stated locally, or it becomes another piece of comfortable nonsense.

Presence is the capacity to be there — to remain a subject in relation to a world that still hangs together. It can fail.

Ernesto De Martino gave the failure a name: the crisis of presence. A culture supplies an order of names, roles and values that lets a person trust the ground. When that order stops holding, what goes is not only confidence but the ability to constitute a world of objects at all. Cultures know this and build against it — ritual, technique, the whole magical-religious apparatus, understood not as superstition but as machinery for defending presence.

Federico Divino reads early Buddhism against that framework and finds the terms already sitting there. Loka, the world as an order of divisions and designations. Dukkha as intrinsic to that order rather than an accident occurring inside it. And the ascetic who walks to its edge: the world-ender who is also, and not coincidentally, the healer — whose knowledge of illness comes from the forest, from outside the village where the naming happens.

The asymmetry is the interesting part. For De Martino the end of the world is a collapse to be defended against. In the Buddhist reading it is undertaken deliberately, as a technique. So which of the two is happening to someone who has accepted that the future will not behave? That question is the whole thread, and I do not think it resolves.

The design consequence

If presence is something a culture equips, then a garden is one of those devices: an unglamorous technology by which a person stays a person on a bad day. That is a design brief, not a metaphor, and it can be specified.

After Divino, F. (2023). Reaching the End of the World: An Anthropological Reading of Early Buddhist Medicine and Ascetic Practices. Religions 14:249 — and De Martino on presence and cultural apocalypse.

The surprise is not always in the weather. Often it is in us.

Kelder and colleagues make a distinction worth stealing outright: you can be surprised by the hazard, or by the impact. For the hazard, they set out four complementary lines of evidence for events outside the record — extreme-value statistics on observations; past events recovered from rescued archives, documents, proxies and oral history; physically plausible event storylines; and large model ensembles. Each is weak alone. Together they can establish that something is possible before it has happened even once.

For the impact, they use an adaptation pyramid. Reactive at the top: warning and response. Incremental in the middle: upgrading what already exists. Transformative at the base: changing the structures that produce the vulnerability in the first place. Thin at the base and the pyramid is unstable — everything ends up resting on warning people faster.

Where a garden actually sits

Mostly incremental. Sometimes reactive, if it is the cool place a neighbourhood is told to go on a red-alert day. Occasionally transformative, if building it changes who gets to decide what a street is for. Being honest about which one is happening in a given project is most of the work, and most of the failure.

The disbelief problem

One finding stays with me: warnings about conditions outside experience are often not believed. People treated a flood as impossible because it arrived out of season. That is not a data problem. It is a presence problem — and it is exactly where the climate science and the anthropology meet.

Measuring it

From the arts-based side comes a useful caution about impact language: ripples, resonance, effects that unfold on their own schedule, held against metrics that only count what arrives on time and in the agreed unit. Both are needed. Neither should be allowed to stand in for the other, and the second one usually tries.

After Kelder, T. et al. (2025). How to stop being surprised by unprecedented weather. Nature Communications 16:2382 — and Dutra Gonçalves, R. & Toiviainen, A. (2026). Navigating Uncertainty. Research in Arts and Education 2/2026.

Art direction and research, based in Reykjavík. Anthropology, futures work, and a long apprenticeship in showing people a landscape.

Currently

• NowArt director, state institution — Reykjavík
• OngoingIndependent research: collapse discourse, futurity, climate communication

Previously

• —Field research: environmental anthropology, tourism studies, embodied and microphenomenological methods
• —Tour guide, Iceland — the original training in narrating a place to strangers

Work

• —Foresight instruments: causal layered analysis, scenario tools
• —Educational design for adolescents facing an unstable future
• —Documentary and data-driven visual work

Replace these entries with your own — this section is a scaffold, not a claim.

Open to collaboration, commissions, teaching, and disagreement made in good faith.

hello@example.is

Reykjavík, Iceland · replies are slow and considered rather than fast

What is useful to send

A specific site. A specific objection. Material that complicates the premise. What is less useful: general enthusiasm about the topic, and invitations to be optimistic on a panel.

This floor is not the bottom. It is where the light comes in.

The garden is a landing, not a rescue. It was built out of what was to hand, on a staircase that keeps going down past the door in the far wall. Treating it as an arrival is the mistake the whole project is about.

There is a Japanese word, ma, for the held interval. The last flight of stairs is the useful part of the descent, not the thing you get through in order to reach the useful part. Not knowing is where the work happens.