Nothing Melts

Walk on snow that fell an hour ago and you go straight through it. Walk on the same snow two days later and it carries you.

Nothing was added in between. Nobody packed it down, no water was poured on, nothing froze and thawed. The same crystals are lying in the same place in the same quantity, and at some point they stopped being a powder and became a surface.

A pile of separate things became a material, and the only ingredient was time.

The name for what happened is sintering, and it is not specific to snow. It is how a ceramic mug is made, how metal powder becomes a machine part, how the icy grains on a comet bind into a crust. In each case a heap of loose particles turns into a solid without anyone melting it.

That last clause is the whole of this zine. The obvious way to make a heap of separate things into one thing is to melt them, so they run together and cool as a single mass. It works, and it is not what happens here.

Sintering happens below the melting point. The grains do not liquefy, do not flow together, and do not cool into a single new mass. They stay solid the entire time.

And they stay themselves. After sintering, a snowpack is not one continuous piece of ice. It is the same crystals it was before, still individual, still with their own boundaries, now joined at the points where they meet. You can section it and count them.

The joining does not cost them their edges. Nothing is dissolved into anything, and afterwards you can still say where each one is.

Held honestly, because the title overclaims slightly and we would rather say so than be caught at it. Ice is not simply solid right up to 0 °C. Its surface carries a quasi-liquid premelted layer below the bulk melting temperature — the thing that makes ice slippery when it has no business being slippery — and that layer can help grains bond without bulk melting occurring. Its onset temperature, its thickness, and its actual relationship to ordinary liquid water have been argued over for more than a century and are not settled. So: nothing melts in bulk, no grain becomes a drop, and no crystal loses its identity — but the surface where the joining happens is not a simple solid either. The interesting part of the claim survives; the tidy version of it does not.

Which is worth sitting with before the mechanism arrives. There is a way for separate things to become a strong single structure that does not require any of them to stop being separate things.

And this zine’s title is not our observation. Leanne Betasamosake Simpson makes the point herself, in a parenthesis, describing a snowflake forming bonds “at temperatures below zero (this is not a melting process) with its neighbouring snowflakes or crystals to create the fabric of a snowpack.” See spread ten, which is about whose reading this is.

Where two grains touch, a small bridge grows between them. It is called a neck, and the question worth asking is why it forms there.

Ice sublimes: molecules leave the surface into the air as vapour, and arrive back from it, all the time. What matters is that the rate depends on the shape of the surface they are leaving. A convex surface — the outside of a grain — holds its molecules slightly less tightly and gives up more vapour. A concave surface — the little hollow where two grains meet — holds them slightly more tightly and gives up less.

So there is a standing imbalance. Vapour leaves the rounded outsides of the grains and preferentially condenses in the crevice between them, simply because that is where the vapour pressure is lowest. The neck fills in because it is a neck.

The connection is not built at the connection point by coincidence. It is built there because a joint is, geometrically, the place material goes.

And the bridge is self-reinforcing while it is small: the tighter the curve of the hollow, the stronger the pull, so the earliest and most awkward stage of a bond is the one the physics favours most.

Vapour diffusion is the main route in ordinary snow. It is not the only one — surface diffusion matters for very small necks, and other transport paths contribute — but the direction is the same whichever mechanism carries it. Material goes to the contact.

Nothing arrives from outside. No binder is added, no glue, no external supply. Every molecule in the bond came off the surface of one of the grains it is joining.

Which means the grains are very slightly diminished by it. Material leaves their rounded outsides and goes into the bridge, so as the neck thickens the grains lose a little of their own convexity to it. The joint is paid for out of the substance of the things being joined, and by nobody else.

The connection is not free, and it is not subsidised. It is made of them.

It is worth being exact about the scale of that cost, because it would be easy to make it sound tragic and it is not. The grains are not consumed. What they give up is a small amount of surface, and what they get is a structure that holds. In the accounting of the whole snowpack the trade is enormously favourable: a modest redistribution of material buys a very large increase in strength.

But it is a redistribution and not a gift, and a piece about bonding that skipped this would be describing a machine that runs on nothing. The bridge is not extra. It is some of them, relocated.

A neck forms at contact. Not near contact. Grains that are close but not touching do not bond, however long you leave them and however many of them there are.

The rate is controlled by things that are all about the contact itself: how hard the grains are pressed together, and what shape they are where they meet. Angular grains with real contact area bond differently from rounded ones resting against each other at a point.

Proximity is not contact. A heap of grains in the same box is not a snowpack, and time alone does not convert one into the other.

This is the least poetic fact in the zine and possibly the most useful one. The mechanism on the previous two spreads is real, reliable and free — and it is entirely conditional on an actual meeting having taken place. The physics does not reach across gaps. It has no mechanism for that at all.

So a structure of this kind cannot be assembled by putting the components in the same place. Something has to bring them into contact, and hold them there, before any of the rest of it can begin.

Sintering is strongly temperature-dependent, and not gently so. Close to the melting point necks grow very rapidly. In deep cold the same process runs so slowly that snow can stay powder for a long time.

Nothing about the grains has changed between those two cases. They are the same crystals with the same shapes in the same contact. What differs is the amount of thermal energy available to move molecules from surfaces into vapour and back down into the hollows.

Whether a heap of grains becomes a structure is not only a question about the grains. It is a question about the conditions they are in.

This is the spread where the physics stops being neutral and starts being a claim about environments. The capacity to bond is not a property the grains either have or lack. It is a property of grains-in-a-temperature, and the same snow that will not consolidate at −30 °C will knit within hours near zero.

Which sets a limit on how much any account of bonding can ask of the things being bonded. You cannot get a cold snowpack to sinter by describing the benefits of sintering to it.

A sintered snowpack does something none of its grains can do alone: it distributes a load.

Stand on loose snow and your weight goes into whichever crystals are directly beneath your boot, which have nothing to pass it to, so they move and you sink. Stand on sintered snow and the force travels outward through the necks into grains that are nowhere near your foot. The structure spreads the load across members that are not carrying you directly.

That is the entire benefit, and it is worth stating in exactly those terms rather than as strength in the abstract. The snowpack is not made of tougher crystals afterwards. Each individual grain is as fragile as it ever was. What changed is that no grain has to bear the whole thing by itself.

Nothing got stronger. The load got shared, which turns out to be the same thing from the outside and a completely different thing from the inside.

It also becomes far harder to disturb. A single grain can be flicked away; a grain with six necks cannot be removed without breaking six bonds, and the force required goes up accordingly. Resistance to collapse is not hardness. It is connectedness.

Everything so far has been an argument for bonding. Here is the other half, and it is not a small qualification. Sintering is also how a slab is made.

A slab avalanche needs a particular arrangement: a cohesive layer of snow sitting on top of a weaker one. The cohesive layer is cohesive because its grains have sintered together. And when a fracture starts in the weak layer underneath, that cohesion is exactly what lets the break run — because the slab moves as one connected object rather than as loose grains that would simply subside.

The same connectedness that carries your weight is what allows a crack to cross the whole of it at once.

Both halves of that sentence describe the same necks. There is no version of the structure that has the load-sharing and not the propagation; they are one property seen from two sides, exactly as the depth of a well is both the holding and the cost of leaving.

And note where the failure actually is. The slab is not the problem. The slab is doing precisely what a bonded structure does. The problem is the weak layer underneath it, which the bonding above cannot reach, cannot repair, and cannot compensate for. A perfectly sintered slab on bad ground is a perfectly sintered slab on bad ground.

Bonding to each other does not fix what you are all standing on. That is not an argument against bonding. It is a statement about what bonding is for and what it cannot be asked to do, and any account of mutual support that leaves it out is selling something.

Everything up to here is snow physics, and snow physics does not mean anything on its own. The move from sintering to coalition was made by somebody, and it matters a great deal who.

It is Leanne Betasamosake Simpson’s, in Theory of Water: Nishnaabe Maps to the Times Ahead, in a chapter called simply “Sintering.” She is Michi Saagiig Nishnaabeg, and the book is what its subtitle says — Nishnaabe maps, not a general-purpose metaphor kit. Helen Edgar then applied that reading to neurodivergent community in Sintering: Neurodivergent Community Building, which is where this zine came from. The order is her framework, Helen’s application, our physics, and the physics is the part that was already lying about in a review paper.

She did not get it from a review paper. She got it from grooming a ski trail. The chapter is written from the sled, in the early morning: “I wasn’t expecting the snow to teach me how to live in the world.”

Her definition, which is the sentence this whole zine has been walking around:

“Sintering is a joining. It is a communal transformation that creates a fabric of former snowflakes bonded to each other.”Leanne Betasamosake Simpson, Theory of Water, ch. 4

And then the two lines that do the work our spreads three and eight were reaching for, and do it better: “the first thing a snowflake does when it lands from the skyworld is to join bonds, actual physical bonds, with its neighbours” — and “It weaves itself into its environment, and it does so in a way that doesn’t destroy its neighbours.”

Her conclusion is not a metaphor and should not be softened into one: “The snow is telling me that sintering is how Michi Saagiig Nishnaabeg make worlds, how we weave ourselves into the land without destroying it.” She asks whether it could ground a method of solidarity, “strengthening and renewing connections across communities of struggle towards new constellations of co-resistance.”

And she names the cost more honestly than we did. Two sentences, and the second is the one to sit with: “Sintering is slow deformation.” And: “Sintering in the forest is easy for me; sintering with people I’m currently sharing the planet with, much more difficult.”

What is ours, so that it is not mistaken for hers. Spread nine — the slab, the weak layer, and the argument that cohesion is also what lets a fracture propagate — is not in her chapter and is not her claim. She is describing what bonding builds; the avalanche is our addition, and any objection to it is an objection to us. Also credited: Helen’s essay draws on Jorn Bettin, Trust in Human Scale (2024), on being trusted only to the extent that we comply.

A piece that takes an Indigenous writer’s framework and runs physics over it can go wrong in several directions at once. So, plainly: