From the Normalbaum to the label: being able to unfold the simplification

I have already written here about James C. Scott and his Normalbaum.

Let me simply recall the idea.

To make a complex reality understandable and manageable, you sometimes have to make it legible. A forest is an extraordinarily diverse reality. Yet to be able to manage it, the forester has to describe it through a handful of characteristics: species, sizes, ages, volumes, categories.

The Normalbaum is therefore not "the real tree".

It is a useful representation of the tree.

But this story has an ending, and I have already told it here: the standardised forest of Prussia, replanted in rows of a single species to match its own representation, thrived for one generation. The second died. Not because the foresters were incompetent — they were the best of their time — but because, by dint of managing the map, they had ended up replacing the territory. That is the story told in What Our Registry Will Never Be Able to Write.

It is this double lesson that interests me today.

Simplification is necessary.

And simplification, left alone long enough, ends up acting on what it describes.

A food is a story

Take a food product.

Before it is a product, it is a succession of events.

A seed. A plot. A soil. A farmer. Practices. Weather. Decisions. A harvest. Storage. Transport. Possibly processing. Checks. Documents. People who observe, measure, declare or verify.

That story can be extremely long.

And yet, when the product reaches us, we need to be able to make a decision quickly.

That is precisely what categories are for.

The label: a technology of legibility

The organic label is an excellent example — it is even one of the finest successes of legibility of the last fifty years.

It turns a complex reality into information that can be understood at once.

Organic.

In a single word, a considerable amount of information becomes accessible to the consumer.

They do not need to know the regulation in detail. They do not need to consult the certification file. They do not need to reconstruct the history of the plot.

The label does its job: it makes complex information legible. And that is an essential function.

The same goes for many other categories used in food: geographical origin, appellation, production method, certification, product category, regulatory statement.

These categories are not reality. They are interfaces with reality.

And we need them.

What the forest really teaches us

Let us go back to Prussia for a moment.

What killed the standardised forest was not the representation itself. It was the moment the representation became the objective.

When a reading grid succeeds, reality starts to organise itself around it. It is a general phenomenon, one that targets no one and spares nothing. We have seen it everywhere.

Fruit varieties selected for size grade and for how well they travel — the two things the grid measures — rather than for taste, which no box asked about.

Teaching that ends up preparing for the exam rather than for the discipline the exam was meant to measure.

Indicators that, as soon as they are used to decide, slowly stop describing.

This is no label's fault. It is a property of every system of legibility — regulations, specifications, purchasing grids, scores. And it strikes the best categories first, precisely because they are the ones there is most to gain from reducing to a box.

The first to know this are the people who fill in the boxes.

A producer — organic or conventional, it makes no difference — spends a growing share of their time fitting their work into grids: declarations, specifications, size grades, supplier listings. They know better than anyone what the boxes do not say. What they do beyond what the box asks. What the box will never be able to see.

The risk for a label, then, is not that it is a simplification.

It is that it becomes a box you tick instead of a story you tell.

And no label can ward off that risk on its own, because it does not come from the label. It comes from not being able to look behind it.

What the interface abbreviates

The fact that a category is useful does not mean that it answers every question.

"Organic" answers certain questions.

It does not necessarily answer all those one might ask about a product.

And that is perfectly normal.

You may want to know its origin. Or the plot. Or the harvest date. Or the storage conditions. Or the processing it has undergone. Or the checks carried out. Or the data that make it possible to support this or that claim.

At that point, the label is not inadequate.

The question has simply become a different one.

We have moved from:

"Which category does this product belong to?"

to:

"What is the story that lets us understand this product?"

So the problem is not the label

This is an important distinction.

It would be absurd to blame a label for not containing a product's whole story. That is not its function.

A road sign is not the road. A map is not the territory. A label is not the full body of information that led to it being awarded.

But the Prussian forest taught us that a map that can no longer be checked against the territory ends up replacing it.

The question therefore becomes: how do we keep the maps — we need them — while keeping the path to the territory?

From traceability of the product to traceability of the claim

Food traceability is often thought of as a question of flows:

Where does this product come from? Where did it go? Which stages did it pass through?

That is essential.

But there is another, complementary question:

Why can we claim this about this product?

This question takes us from the product to the information.

Saying:

"This product is organic"

is a claim. Saying:

"This product comes from such-and-such a farm"

is another. Saying:

"This raw material was checked on such-and-such a date"

is yet another.

Behind each of these sentences lies a set of rules, events, documents, observations or declarations. It is that path that interests us.

Proof is a path

A certificate is not the event it certifies. A document is not the event it describes. A declaration is not an observation. A measurement is not reality itself.

They are elements that make it possible to build, support or verify a claim.

That is why we like this wording, which gives its title to another article:

Proof is a path.

That path may be short.

It may be long. It may be fully documented. It may include areas of uncertainty. It may even bring contradictory information to light.

But in every case, it lets us understand not only what is being claimed, but what that claim rests on.

Unfolding does not undo categories — it defends them

This is perhaps where the parallel with Scott becomes most useful. What the standardised forest lacked was not a better grid. It was the possibility of going back to look at the forest.

A category that can be unfolded — whose origin, rules and supporting elements can be seen — no longer runs the same risk of substituting itself for reality. It remains what it should be: an interface.

And that shifts the balance of power in favour of honest categories.

A label you can unfold is worth more than a label you have to believe.

Those who have a real story behind their box have everything to gain from it being open to view. Organic does not need to be set against conventional to be documented. A conventional farm does not need to be set against an organic farm to be documented. A geographical indication does not need to be set against another way of producing to be documented.

In every case, the same question still applies:

What makes it possible to support this claim?

And in every case, the one who actually does the work is the one this question favours.

It also changes how we talk about the farming model

Farming debates love big categories.

Organic or conventional. Local or imported. Small or large. Intensive or extensive. Traditional or technological.

These categories are useful for debate. But they quickly fall short when it comes to understanding a concrete farming reality. Two farms in the same category can operate very differently. Two products carrying the same label can have very different histories. And two products that do not carry the same label can share certain characteristics that the category does not let us see.

That does not make categories useless. It simply means they must not be mistaken for the whole of reality — and that a debate that only pits boxes against each other ends up, like the forest, no longer talking about the territory.

The next step: unfolding

For a long time, we mostly sought to make information simpler. That was necessary.

A label. A logo. A certificate. A QR code. A product sheet. A database.

Each step made part of reality more easily accessible. The next step could be different.

It is no longer just a matter of simplifying information.

It is a matter of making it possible to unfold the simplification.

See the label. Then see what it means. Then see the elements that support it. Then trace back, when necessary, to the events and the people at the origin of those elements.

Not to ask everyone to read everything. Quite the opposite.

So that everyone can choose how far they want to go.

The consumer can stay at the level of the label. The professional may want to go further. The auditor further still. The buyer may need a specific piece of information.

And tomorrow, an AI may also need to know not only that a piece of information exists, but where it comes from.

This applies to us too

It needs to be said plainly, because Scott's lesson does not stop at other people's labels.

VeraTrace, too, produces simplifications. Scores. Indicators. Badges. Each of them is a small Normalbaum: a useful representation, and therefore exposed to the same risk as every useful representation.

The rule we set ourselves follows directly from everything above: everything we display must be able to be unfolded. A score whose construction cannot be seen is worth no more than a ticked box — ours included.

That is where VeraTrace wants to stand

VeraTrace is not trying to replace labels.

Nor certifications. Nor inspections. Nor producers. Nor regulations.

It is trying to make the relationships between them more legible. To turn a piece of information into a path you can navigate.

To make it possible to move from:

"Here is what is claimed"

to:

"Here is why this claim exists."

It is a simple difference.

But it becomes important in a world where we produce and consume more and more information about products.

The Normalbaum reminds us that every representation of reality is a simplification — and that simplification, once we can no longer look behind it, ends up reshaping what it described.

The label shows us that this simplification can be extremely useful. Traceability may now give us the possibility of doing something the Prussian foresters could not do:

not abolishing the simplification, but being able to unfold it. Because a good information system should not force us to know everything. It should simply let us, when it matters, know where to look to find out more.