Selling Wine With Bottles
A reply to John Perry Barlow, thirty-two years on
St. Louis — November 13, 2026
John Perry Barlow asked the right question and then, I will argue, answered it with the wrong bottle.
Buried in the middle of his 1994 essay — the one that told a generation the bottle was gone for good — he paused to ask, in plain terms, the question everything since has been an attempt to dodge:
“What technologies will be useful in creating new virtual bottles to replace the old physical ones?”
This essay is an answer, thirty-two years late. We found the bottle. It was never a container. The bottle is the information context — and context begins with identity.
The riddle, re-read
Barlow opened with Jefferson and a riddle:
“The riddle is this: if our property can be infinitely reproduced and instantaneously distributed all over the planet without cost, without our knowledge, without its even leaving our possession, how can we protect it? How are we going to get paid for the work we do with our minds?”
Notice what the riddle assumes: that protection means preventing the copy. Every failed answer since 1994 — the lawsuits, the shutdowns, the DRM, the fines — has inherited that assumption and broken on the same rock. The copy cannot be prevented, and should not be. Bits copy identically; that is their nature, and no law has ever repealed a nature.
But protection has a second meaning the riddle never considers: not containment of the work, but title to it. You do not protect a deed by making paper uncopiable. You protect it by keeping one authoritative record of whose it is. Barlow’s question — how are we going to get paid for the work we do with our minds? — has an answer that never required stopping a single copy: you get paid at the event, by attestation, because every legitimate copy carries its context with it. The wine flows free. The bottle — the record of whose it is, when it was made, and on what terms it may be used — travels with every pour.
Two jobs of a bottle
Barlow’s most famous sentence is also his most careful, and it deserves to be quoted whole:
“One didn’t get paid for ideas but for the ability to deliver them into reality. For all practical purposes, the value was in the conveyance and not the thought conveyed. In other words, the bottle was protected, not the wine.”
True — and incomplete. The physical bottle did two jobs, and Barlow, watching the first one die, concluded the bottle itself was finished. Job one was containment: scarcity, the shrink-wrap, the thing you could hold and therefore sell. Digitization abolished it, permanently, and good riddance.
Job two was attestation. Even in the old physical world, the bottle’s deepest magic was never the glass; it was the binding of contents to origin. A great wine is not sold as fermented grape juice — chemically, that is all it is — but as the produce of a named estate, a named slope, a named year, sealed at the source so the claim and the contents could never be casually separated. The appellation systems of Europe are registries of origin centuries older than copyright. The value was never only in the conveyance. It was in the provenance the conveyance carried.
Digitization killed containment and left attestation homeless. That homelessness — not copying — is the disease of the digital age. We spent thirty years mourning the wrong function of the bottle.
Shannon’s bracket
Underneath Barlow stands an older and greater engineer. In 1948 Claude Shannon founded information theory by deliberately setting meaning aside:
“Frequently the messages have meaning; that is they refer to or are correlated according to some system with certain physical or conceptual entities. These semantic aspects of communication are irrelevant to the engineering problem.”
Shannon was honest: it was a bracket, a scope limit, true for the problem he was solving — how to move signals faithfully through noise. But an industry took the bracket for a metaphysics. If the channel treats all bits alike, the reasoning went, then bits are alike — fungible, ownerless, poolable — and whoever aggregates the most of them owns the age. Every platform, every data broker, every omnibus ledger in which your life sits as a row in someone else’s table is Shannon’s bracket, forgotten on purpose, at scale.
Barlow, to his credit, saw the hole in the abstraction and walked halfway back. “In most cases,” he wrote, “we assign value to information based on its meaningfulness.” And more precisely still:
“In fact, information, in the economic sense of the word, consists of data which have been passed through a particular human mind and found meaningful within that mental context.”
Within that mental context. Barlow located value in context and then went looking for the bottle in scarcity, performance, and relationship — everywhere except the place his own sentence pointed. So here is the rebuttal, to Shannon’s followers and to Barlow alike, stated as flatly as I can state it:
All information is ones and zeros, but some ones and zeros are more valuable than others.
The differential is not entropy. It is context: whose it is, who made it, when, under what attestation, in what chain of custody. Two hundred fifty-six bits of a private key and two hundred fifty-six bits of static are identical to Shannon’s mathematics and universes apart in the world, because one string has a person standing behind it and the other has no one. A bit with a root is a deed. The same bit without one is weather.
Say it as a rule of evidence for the digital age: if the information is stripped of the context, it is not legitimate. Not necessarily false — illegitimate: no standing to be believed, traded, or relied on, because no one stands behind it and nothing anchors where it has been.
This is also the whole story of Napster, told correctly. When we realized we could copy the bits identically, it seemed to be a license. It was not — it was a confusion of transmission with title. Napster did not distribute music; it distributed orphans: perfect copies severed from their context, unable to say who made them or where a penny of reward should flow. And the lesson worked both ways. The generation that copied the labels’ catalogs taught the labels’ successors to copy the generation — faces, voices, friendships, lives — under the same forged license. The users pirated the songs. The platforms pirated the users.
“It doesn’t really work. Honest.”
Barlow was a working songwriter, an ASCAP member, and his testimony about the old world’s accounting remains the most damning paragraph in the essay:
“But, as a member of ASCAP, I can assure you this is not a model which we should emulate. The monitoring methods are wildly approximate. There is no parallel system of accounting in the revenue stream. It doesn’t really work. Honest.”
Thirty years later it still doesn’t: royalties pool in black boxes, unmatched money sits with collecting machinery because the books cannot say whose it is, and the creator is settled-up against a ledger he may not read. But the conclusion Barlow drew — that accounting for performance was hopeless, so creators must retreat to live performance and patronage — was a 1994 conclusion, drawn when monitoring really was approximate. It no longer is. A performance is now the cheapest thing in the world to attest: a few bytes — the work, the moment, the count. Small data, lots of it. Let every performance be an attested event at the work’s root, and payment stops being a settlement and becomes a reflex: the event is the accounting. The black box dies of arithmetic.
Many patrons — and the deed
Barlow foresaw the shape of the new patronage with real precision:
“Before the industrialization of creation, writers, composers, artists, and the like produced their products in the private service of patrons. Without objects to distribute in a mass market, creative people will return to a condition somewhat like this, except that they will serve many patrons, rather than one.”
He was right about the many patrons and wrong about the condition they would serve in. The old patronage’s price was the deed: Haydn’s 1761 contract made his compositions his prince’s property, uncopyable and unsellable by the man who wrote them — and when the clause fell away in 1779, the same talent became, within a decade, the most famous composer in Europe. The talent never changed. The title did. What Barlow’s future kept from the old world was the patron’s ownership; what it should have kept was only his purse.
So amend his sentence by the one distinction that matters: everyone is a patron — and no one is a lord. Funding and owning, fused for five hundred years, come apart. The patron pays, commissions, enjoys; the title never travels with the money. Intellectual property in the new bottle is rooted at its creator, and assignable, and sellable — real ownership, alienable by its owner’s hand alone, because a root that cannot move is a cage and a license that severs the name is a theft. This is why the bottle matters, finally, in one sentence: because it respects the creator. A copy whose context can never be peeled off is respect made a property of the medium itself.
Crypto bottling, completed
Barlow even named the material of the new bottle, in a section he himself titled Crypto Bottling:
“Cryptography, as I’ve said perhaps too many times, is the ‘material’ from which the walls, boundaries — and bottles — of Cyberspace will be fashioned.”
“But cryptography will not be used simply for making locks. It is also at the heart of both digital signatures and the afore-mentioned digital cash, both of which I believe will be central to the future protection of intellectual property.”
He pointed at the right tools and reached for the wrong handles. Signatures, in the world that got built, were put in service of institutions: platform accounts, custodial wallets, locks on other people’s doors — until, in the largest crypto theft in history, thieves walked through an exchange’s back room using the house’s own machinery while its vaults stood perfect and unpicked, and the company’s truest sentence was that no private key had ever been compromised. The keys were fine. The authority was unrooted — held by services and laptops instead of persons.
The bottle Barlow asked for is a signature with a person inside it: a verified human root, minted in presence — the handshake as protocol — from which works, assets, credentials, and agents all hang as leaves. Trade between such roots is an old, honorable form wearing new mathematics: the structured, agreed transaction between known entities that commerce once ran through EDI envelopes — an outer envelope naming the parties, an inner one carrying the goods, nothing existing until both parties agree. Barlow hoped “the protections which we will develop will rely far more on ethics and technology than on law.” He was right, and the ethics are older than the technology: honest scales; the worker’s wage not held overnight; no value resting in transit in a stranger’s name. The technology, at last, can keep the ethics.
Flat statements, amended
Barlow closed with flat statements he believed would not look silly in fifty years. Thirty-two have passed. Here are his, and mine, side by side.
His: “In the absence of the old containers, almost everything we think we know about intellectual property is wrong.”
Mine: In the presence of the new container, almost everything he feared about intellectual property is fixable. Everyone is focused on the symptoms, not the solutions. What is wrong is fixable.
His: “The economy of the future will be based on relationship rather than possession.”
Mine: The economy of the future will be based on relationship and possession — patrons many, title one, the creator’s — because a relationship without title is patronage, and we have tried patronage.
And the last one is the whole reply:
Barlow: wine without bottles.
We found the bottle. It is identity — the context of the information, bound to its maker, traveling with every copy, answerable to its owner alone.
Jefferson, whose words opened Barlow’s essay, said that he who lights his taper at mine receives light without darkening me. Barlow read that as the end of property. Read it again. The light spreads free, exactly as Jefferson hoped — ideas to everyone, regardless of station. But the candle keeps its owner, and so does the house it stands in. The wine was never the problem. It only needed its bottle back.
It’s your data. Stop giving it away.
A Testimony
I did not come to this argument in a library. Before my newspaper had published a word online, we built an internet service from scratch — sold the disks in supermarkets, like AOL, so people could get to the internet and find us, with a Netscape browser to open the door. And when we did put the St. Louis Post-Dispatch online, we were among the first newspapers in America to do it. I was the CIO/CTO, and my goal was never to sell content. It was to build a community around the news. The WELL was my model; the MIT Media Lab was my influence. The news as a campfire. The readers gathered around it, talking back.
I think we did well. People came, and stayed, and belonged.
Then it was plastered as a paywall. The wall went up around the gathering, to charge for what was inside. And the value walked away. It moved to social media — no context, only shouting. The owners kept the container. The wine was gone.
So I have seen both failures with my own eyes. Wine with no bottle — scattered, copied, orphaned. And a bottle with no wine — sealed, priced, and empty. There is a third way, and it is the argument of this essay: keep the wine free, keep the gathering open, and give every drop its maker’s name.
— Richard Truex, St. Louis