Recipes

    Today

    Nine pages to get from a perforated clay sieve in Neolithic Poland to here. This one is about where the craft actually stands, what’s coming that might change it, and one question the frontier asks that I don’t especially enjoy answering.

    Both extremes at once, and that’s new

    The honest description of cheese in 2026 is that it’s more industrial and more artisanal than it has ever been, simultaneously.

    The commodity end is the direct descendant of everything on the industrialization page: enormous, standardised, cheap, safe, consistent, and mostly cheddar-shaped. That system feeds an extraordinary number of people, and the snobbery about it usually forgets that consistency was a genuine problem somebody had to solve.

    At the same time there is more good cheese available to an ordinary person than at any point in history. That’s not nostalgia — it’s a fact about distribution. A cheese counter in a mid-sized American city now carries things that a French duke couldn’t have eaten in 1700, because they’d have spoiled before they arrived.

    And the third thing, the one this whole section exists for: you can make it at home. Cultures ship in the post. The books are written. Somebody will sell you rennet, a press and a mould, and it costs less than most hobbies. The knowledge that Harding published in 1860 and Williams refused to patent is now sitting free on the internet, which is arguably the truest continuation of what either of them was doing.

    The fight over names is still running

    The naming thread that ran through monasteries, the Alps and England has not resolved. It has become trade policy.

    In July 2022 the European Court of Justice decided Case C-159/20. Feta has been a PDO since 2002, restricted to Greece and Cyprus. Denmark had allowed Danish dairies to keep producing cheese labelled “Feta” — but only for export outside the EU — arguing the rules bound them within the bloc and no further. The court disagreed: Denmark had failed its obligations under Regulation 1151/2012 by not stopping production of “Feta” even when destined for third countries.

    That’s the 1411 Roquefort charter grown up and gone global. A protected name is now enforceable against your own producers on behalf of another country’s product, for goods you’re shipping somewhere else entirely. And the transatlantic version of the argument — whether “parmesan,” “gruyere,” “feta” are protected origins or generic English words — is unresolved and genuinely hard, which is why nobody has settled it.

    🔴 The frontier: cheese without a cow

    Here is what’s actually new, and it is a straight continuation of a story this wing already told.

    On the industrialization page: in 1990 the FDA approved fermentation-produced chymosin — the calf’s rennet enzyme, made by microorganisms instead of a calf. The first bioengineered product in the US food supply, and it’s in most American cheese now.

    That was the enzyme without the calf. The frontier is the protein without the cow.

    Precision fermentation puts the gene for a milk protein into a microorganism and ferments it in a tank. Perfect Day got an FDA letter of no objection for animal-free whey protein in March 2020. Then New Culture, in California, self-affirmed GRAS for animal-free casein — the world’s first, clearing it for commercial sale in the US, and aimed squarely at mozzarella. Berlin’s Formo received a “no questions” letter from the FDA for its precision-fermentation casein in June 2026.

    The detail that matters is in the GRAS determination: the casein is bioidentical to the cow’s — same identity, same macronutritional profile. Not an analogue. Not a substitute that behaves similarly. The same molecule.

    The question I don’t enjoy

    I wrote on my bookshelf that nut-based vegan “cheese” isn’t cheese by the definition this section uses — no milk, therefore no casein, therefore none of the chemistry on the family tree applies. I stand by that, and I said it was a point about vocabulary rather than quality.

    But a definition you only apply when you like the answer isn’t a definition.

    Precision-fermented casein is casein. Add a coagulant and it does what casein does — the micelles destabilise, the network forms, you can cut it and cook it and press it and age it. Every mechanism on that diagram applies, because the molecule the diagram is about is present.

    So by my own tree, cheese made from precision-fermented casein is cheese. Not cheese-like. Cheese. I find that uncomfortable, and I’ve thought about the ways out, and they don’t hold up:

    • “It didn’t come from an animal.” Neither does the rennet in most American cheese, and has since 1990. That ship sailed with the last generation’s argument.
    • “There’s no terroir, no herd, no season.” True, and that’s a real loss — but it’s a loss of the things that make a cheese particular, not of the things that make it cheese. A block of commodity cheddar has no terroir either, and nobody disputes its category.
    • “It’s not traditional.” The tradition in this section includes a Roman recommending fig sap, a Swiss immigrant reverse-engineering Limburger in Wisconsin, and a Bronze Age Chinese kefir culture. Novelty has never been disqualifying.

    The honest position is that my definition is about chemistry, and it gives me an answer I didn’t order. That’s what a real definition does.

    What I think actually happens

    Predictions are cheap, but the shape of this one follows from the whole wing.

    Precision fermentation will compete where consistency and cost are the point — which is precisely the territory Jesse Williams’ factory system and Kraft’s patent were built to win. If you’re making a protein for a pizza that ships frozen, a tank has real advantages over a herd, and the arguments against it are mostly aesthetic.

    What it will not touch is the other end: a cheese that is a particular thing from a particular place, made by a person, aged in a specific cave for a specific length of time. That is the least substitutable product in food, because its whole value is that it couldn’t have come from anywhere else. Roquefort has survived the industrial revolution, two world wars, refrigeration, global shipping, and Danish feta. It’ll survive a bioreactor.

    The middle is where it gets interesting, and I don’t know the answer.

    What the history actually says

    Ten pages, and one thing runs through all of them.

    Milk arrives faster than it can be drunk, and it spoils. That is the entire problem, and every era solved it with what it had. A Neolithic sieve. A Roman with a fig tree. Monks who could afford to wait three years because the monastery outlived them. Jura farmers who invented a cooperative and a ledger because one farm couldn’t fill a wheel. Alpine makers whose wheels grew to fit a mountain path and then a customs rule. An Englishman who published his method. An American who refused to patent his. A state that centralised the whole thing and nearly killed the variety. People who put it back on purpose. And now a tank in California making the protein directly.

    Same problem. Different tools. Every single time.

    Which is why I don’t think my rebuilt refrigerator is a quaint hobby, exactly. It’s the current domestic answer to a seven-thousand-year-old question, running on a temperature controller and a humidifier and some motors that turn the wheels so I don’t have to open the door. I’m doing what the sieve was doing.

    I just have better instruments, and I know the names of the bacteria.

    That’s the History wing. Next: the science — what’s actually happening in the vat, in proper detail. Being written now.

    Sources

    Comments (0)

    1. Loading…

    Comments are held for moderation — nothing appears until approved.