The Colour You
Could Not Buy

In 301 AD the emperor Diocletian issued an edict fixing maximum prices across the Roman economy, an attempt to arrest an inflation that his currency reforms had failed to contain. Among several hundred listed commodities was purple-dyed silk, capped at 150,000 denarii per pound — a figure that worked out to roughly three pounds of gold. A pound of purple-dyed wool was priced at approximately its own weight in the metal.

The edict is a useful place to begin because it converts a familiar claim into an auditable one. Purple cloth being worth more than its weight in gold is the kind of statement that circulates without evidence, and here the evidence exists in the form of a price schedule published at imperial instruction and carved into stone across the eastern provinces. The Greek historian Theopompus, writing in the fourth century BC, reported that purple fetched its weight in silver. Six centuries later the same commodity was being reckoned against gold at a multiple.

What the schedule does not explain is why an empire would go on to make that same commodity illegal for almost everyone. A good already priced beyond the reach of all but a few hundred people does not obviously require a law. The answer to that question is the more interesting part of the history, and it concerns the difference between a signal that is expensive and a signal that is impossible.

Tyrian purple was extracted from the hypobranchial gland of several Mediterranean molluscs, chiefly Murex brandaris and Murex trunculus. The gland yields a clear secretion that develops colour only through exposure to air and light, passing through yellow and green before settling into the deep red-violet that gave the dye its value. The process resisted every kind of shortcut — the whole sequence depended on biochemical, enzymatic and photochemical reactions that no ancient dyer could have described and every ancient dyer had to control.

The yields are the whole story. David Jacoby's figure, widely used in the scholarship, is that twelve thousand Murex brandaris produce roughly 1.4 grams of pure dye, enough to colour the trim of a single garment and no more. On those numbers, dyeing a full garment consumed snails in the hundreds of thousands. At Sidon, excavation has revealed a mound of discarded murex shells roughly forty metres high, accumulated over centuries of continuous production. The workshops themselves were placed at the edges of cities because the smell of the process was, by every ancient account, intolerable at close range.

What matters economically is that this inefficiency could not be engineered away. Grain yields respond to better technique, and most ancient industries improved gradually as methods spread. Murex yields did not, because the binding constraint was the biology of a wild animal harvested one at a time from a fixed stretch of coastline. The dye was expensive in the way that very few goods are genuinely expensive — as a direct function of the labour and material embedded in it, with no plausible substitute at equivalent quality.

There is a second property that the price alone does not capture, and it may matter more than the scarcity. Most ancient dyes degraded. Plant-based colours faded with sunlight, washing and time, which meant that a dyed garment was a depreciating asset. Tyrian purple behaved differently. Rather than fading on exposure, it deepened, and its resistance to light and washing was remarked on throughout antiquity as one of the qualities distinguishing genuine sea-purple from imitation.

That durability changes the economics of the purchase in a specific way. A garment that holds its colour is not consumption but a store of value, transferable by inheritance and legible as such. Scarcity establishes a high price at the moment of sale; permanence is what allows the object to keep signalling across a generation. A dye this expensive that faded within a season would have been an extravagance rather than an institution, and it is difficult to imagine the language of imperial legitimacy attaching itself to something that visibly wore out.

A signal is informative only to the extent that it is costly to produce, and costly specifically for those who lack the quality being signalled. On that reasoning, Tyrian purple should have been self-enforcing. Almost nobody outside the top of the imperial economy could have afforded it, and no law should have been necessary. Rome, and then Byzantium, legislated anyway.

There are two reasons a state might legislate a good that is already unaffordable, and both were operating. The first is that the price barrier was never as absolute as the headline figures suggest. Cheaper approximations existed throughout antiquity, and a wealthy provincial could acquire something that read as purple at conversational distance. Price excludes the poor. It does not exclude the merely rich, and it does nothing whatever about imitation.

The second is that legal exclusion is cheaper to maintain than economic exclusion. Keeping a signal expensive requires the underlying scarcity to persist — a matter of biology, geography and trade routes, none of which any emperor controlled. Keeping it illegal requires only enforcement, which the state already possessed and was already paying for. A monopoly that also writes the law governing its product is in an unusually comfortable position.

The most sophisticated element of the Byzantine system solved a problem that money cannot solve on its own. Any purchasable signal is, in principle, available to anyone who accumulates enough money — which makes it permanently vulnerable to the merchants, provincial magnates and successful generals whose wealth periodically rivalled the court's. A status marker that can be bought is a poor defence against a class of rivals defined by their capacity to buy things.

The Byzantines addressed this by constructing a chamber in the imperial palace lined with porphyry and reserving it for imperial births. A child born within it was porphyrogennetos — born in the purple — and the term entered the vocabulary of succession as a claim to legitimacy distinct from and superior to the claims of usurpers and generals.

The elegance of the arrangement lies in what it converts. Purple cloth is a commodity, and commodities can in principle be acquired. Birth in a specific room on a specific occasion cannot be acquired at any price, at any later date, by anyone. The empire took a signal grounded in the market and produced from it a second signal grounded in a fact about the past — which is the one category of thing that wealth cannot reach. Born in the purple remains in use in English today, describing a kind of legitimacy that money is understood not to confer, which is precisely the distinction the chamber was built to create.

The penalties attached to unauthorised purple were severe, and the reason is that in a court organised around proximity to a single person, dress was not decorative but jurisdictional. To wear the emperor's colour was to make a claim on the emperor's position, and the distinction between a sartorial error and an act of treason was not one the system had any interest in drawing.

Claimants to the throne understood the grammar well enough that assuming the purple was among the first acts of a rebellion, functioning as a declaration that required no accompanying text and could not be walked back. The system made a colour into a commitment device. Wearing it was irreversible in its meaning, which meant nobody could put it on tentatively, and that in turn made the signal reliable in exactly the way signalling theory requires.

The system's eventual failure followed from its design. A monopoly that restricts demand to a single buyer is also a monopsony, and an industry with one customer has no capacity to absorb disruption. The great dyeworks of Tyre and Sidon had, over centuries, been reorganised around imperial supply. Whatever independent commercial market might once have existed had been legislated away.

Several pressures then converged. The murex populations of the eastern Mediterranean were depleted by centuries of harvesting — a straightforward case of an unmanaged common resource extracted past its capacity to recover. Trade routes shifted and imperial territory contracted, cutting the empire off from coastal sites where the work had to be done close to the water. Cheaper dyes improved enough to approximate the colour at a distance. When Constantinople fell to the Ottomans in 1453 the imperial dyeworks lost their only customer, and the industry did not survive the loss. A monopoly had protected the product so thoroughly that it left the product unable to survive without protection.

The arrangement Byzantium built has not disappeared so much as changed hands. Contemporary luxury does not rely on natural scarcity, since almost nothing sold at a luxury price point is genuinely scarce in the way murex dye was scarce. It relies instead on a legally protected mark, and on a body of trademark and anti-counterfeiting law that the state enforces on the holder's behalf. The structure is recognisable: a signal whose value depends on restricted access, access restricted by law rather than by cost, and enforcement supplied by public authority at public expense for private benefit. What has changed is who holds the monopoly. In Byzantium the state owned the signal and enforced its own claim. In the modern arrangement the state enforces a claim it does not own. The mechanism is the same. Only the beneficiary has moved.

On the night of 28 May 1453, with the Ottoman assault about to break through the Theodosian walls, Constantine XI is recorded by most accounts as having removed his imperial regalia before entering the fighting. He did not put the purple on. He took it off, in order to die as an ordinary soldier among his men, and the decision is the leading explanation for why his body was never identified afterward.

Read economically, the gesture is precise. For eleven centuries the colour had functioned as a signal because an entire apparatus stood behind it: the dyeworks, the sumptuary law, the executions, the palace chamber, and the machinery of a state capable of enforcing who was permitted to be seen in it. In the hours when that apparatus ceased to exist, the signal inverted. What had been the most valuable thing a person could wear became the single most dangerous thing to be found wearing, and its last holder discarded it before walking into the dark.

No status marker carries value on its own. It carries the value of whatever institution is prepared to enforce it, and it holds that value for exactly as long as the institution does.