The most widely believed fact about the Silk Road is that it was closed in 1453 when the Ottomans took Constantinople, forcing Europeans to seek a sea route to the Indies, and it is wrong in at least three separate ways.
The first lesson of this course flagged the sentence as suspicious on its form, because a diffuse relay system running through dozens of polities cannot be switched off by one conquest at one end. This lesson supplies the evidence, replaces the story with a better one, and arrives at an ending date four centuries later than the traditional one.
The story, and where it comes from
The standard chain runs: Constantinople falls in 1453; the Ottomans block or ruinously tax the eastern trade; spice prices in Europe rise; Portugal and Spain are driven to find a sea route; Vasco da Gama reaches Calicut in 1498; Columbus sails west in 1492; the Age of Discovery begins and the overland trade dies.
It is a wonderfully tidy causal chain, it appears in textbooks in every European language, and it has the shape of a good exam answer. Test it clause by clause.
Three ways it fails
Constantinople is not on the road. Look at the map the third and fourth lessons built. The overland route to China runs through Persia, Merv, Bukhara and Samarkand, then the Pamirs and the Tarim. Its western outlets are the Black Sea ports, the Levantine coast at Aleppo and Tripoli, and Alexandria via the Red Sea. Constantinople is a great commercial city, and it is a terminus for Black Sea and Mediterranean shipping rather than a link in any chain to Asia. Taking it changes who taxes one branch of the western end.
The Ottomans wanted the trade. This is the decisive objection. A state that has conquered a commercial crossroads does not destroy the commerce it has just acquired; it taxes it, and it protects the merchants who generate the revenue. The Ottomans took Syria and Egypt in 1516 and 1517, which gave them the Levant and Red Sea routes as well, and they ran them as a revenue system, maintaining caravan roads, building caravanserais and negotiating trade capitulations with Venice and later with France and England. Ottoman customs registers are one of the better sources for sixteenth century trade volumes precisely because the trade was going on.
The numbers refuse to cooperate. If the route had been closed, the spice trade through the Levant should have collapsed and stayed collapsed after 1498. Frederic Lane's work on the Venetian records showed the opposite: after a genuine disruption in the first decades of the sixteenth century, when the Portuguese arrival was new and shocking, the Levant spice trade recovered, and by the 1560s Venice was again importing pepper through Alexandria in quantities comparable to those before da Gama. The Portuguese never had the ships to seal the Red Sea, and Asian and Muslim merchants continued to move spices north through Hormuz, Basra and Suez in competition with the Cape route for another century.
Example. Suppose you wanted to test the closure story properly. What evidence would settle it, and which way does the available evidence point?
The claim is about quantities, so it needs a quantity series: the volume or value of eastern goods arriving in Europe, at the same places, before and after 1453 and again before and after 1498. Two series exist in usable form. Venetian records of pepper and spice imports through Alexandria and Beirut, which Lane and later scholars have worked through, show disruption in the early sixteenth century and recovery by the middle of it. Ottoman customs registers for the Syrian and Egyptian ports show a functioning taxed trade throughout. Both point the same way: no closure in 1453, a real but temporary dislocation around 1500 caused by the Portuguese rather than the Ottomans, and continued competition between the two routes for most of the sixteenth century. The evidence that would support the traditional story, a sustained collapse in eastern goods reaching Europe after 1453, has been looked for and is not there.
Now you. If the Levant route was still working, why did Portugal invest so heavily in finding a way round Africa?
Answer
Because a working route and a route you have no share of are different things. Portugal's problem was not that spices had stopped arriving in Europe but that every stage of their arrival was controlled by somebody else: Muslim merchants in the Indian Ocean, Mamluk and then Ottoman authorities in Egypt and Syria, and Venice in the Mediterranean, each taking a margin and none of them Portuguese. A sea route to the source cut all of them out at once, which is a strategic and commercial motive that has nothing to do with blockage. Portugal also had specific advantages that made the attempt rational for it and not for others, an Atlantic coast, decades of accumulated experience working down Africa, and a crown willing to fund voyages. And it is worth noticing what the Portuguese did on arrival, which was to try to seize control of chokepoints such as Hormuz, Goa and Malacca, that is, to become the tax-takers they had objected to. The motive was not access. It was the margin.
What actually shifted
Discard the closure story and a slower, less dramatic account is left, in which several things moved at once.
The Mongol integration ended. The Ilkhanate dissolved after 1335, the Chagatai khanate split, the Yuan were expelled from China in 1368, and the successor states fought each other over the same corridors their predecessors had guaranteed. Timur, who died in 1405, rebuilt an empire and made Samarkand spectacular, but he did it by conquest and by deporting the craftsmen of other cities, and his state fragmented after him. The single most valuable thing the Mongols supplied, a merchant's ability to cross a continent under one law, was gone and did not return.
China turned away. The Ming inherited hostility to the steppe, restricted foreign trade to a tributary framework, banned private maritime trade for long periods, and after 1433 gave up long-range naval expeditions. A frontier held defensively rather than administered is a frontier over which trade is smuggled rather than taxed.
The Ottoman and Safavid states, both formidable and both prosperous, fought each other repeatedly from 1514 onwards, and the transit route through Iran ran through the war zone. Shah Abbas tried to reroute Iranian silk exports around Ottoman territory by sea through Hormuz, which is a good indication of how much the overland corridor had come to depend on the state of relations between two empires.
Undercut rather than blocked
And the sea kept getting cheaper, which is the argument of the previous lesson repeated at a later date.
Larger hulls, better rigs and better navigation lowered the cost per tonne. The Cape route removed the transhipments that had made the old way expensive: goods travelling from India to Antwerp through the Red Sea had to be unloaded, carried overland, reloaded, taxed and handled at every break, whereas a Portuguese carrack loaded at Goa was unloaded at Lisbon. And after 1600 the Dutch and English East India Companies brought something new to the problem, which was permanent pooled capital. A joint stock company can keep ships, warehouses, factories and agents standing between voyages, absorb the loss of a fleet without dissolving, and finance a trade whose capital turns over once a year, which is precisely the difficulty the monsoon imposes.
Against that, the overland route offered a chain of separate merchants, a dozen tax authorities, and a freight rate forty times higher per unit distance. It was not blocked. It was undercut, and the process took two centuries rather than a year, which is why no dramatic date marks it.
The caravans that did not notice
Here is the fact that should finish the traditional story. Long after the road was supposedly closed, it was carrying tea.
Russia and Qing China settled their frontier and their commerce by the treaties of Nerchinsk in 1689 and Kiakhta in 1727, and the second established a designated trading point at Kiakhta on the Mongolian border. Through it ran one of the great caravan routes of history, carrying Chinese tea north and west across Mongolia and Siberia to Moscow, a journey of roughly 11,000 km taking about sixteen months, with Russian furs, cloth and later manufactures going the other way. By the first half of the nineteenth century Kiakhta was handling a large share of Russia's entire eastern trade, and Russian tea drinking, a national habit to this day, was supplied by camel.
Example. The Kiakhta tea caravans ran roughly 11,000 km in about sixteen months. Compare with the third lesson's caravan arithmetic and say what the comparison shows.
Sixteen months is about 486 days, so the average is km a day including every stop, wait and winter, against the 30 km a day of pure travel used earlier. Nothing had changed. A nineteenth century tea caravan moved at the same speed, by the same means, with the same constraints, as a Tang dynasty silk caravan a thousand years before, and it was commercially viable because tea, like silk, is light, dense in value, durable and wanted at the far end. The comparison shows that the overland trade never became technically obsolete: the freight model that permitted it in the eighth century still permitted it in the nineteenth. What ended it was not that it stopped working but that something else began working far better.
Now you. Tea reached Russia overland by camel while reaching Britain by sea in clippers. What does that pairing tell you about how routes are chosen?
Answer
That the choice is made by geography and relative cost rather than by technology, and that both routes can be right at the same time for different buyers. Britain is a maritime country whose ports are on the ocean, so tea from Canton went by sea because for a British importer the sea leg is nearly free by comparison and there is no land alternative at all. Moscow is 1,500 km from the nearest useful ocean, and Russia had no navy in the Pacific and no easy access to the southern sea lanes, so for a Russian importer the caravan across Mongolia was genuinely the cheaper option even at forty times the sea freight rate per tonne-kilometre. Notice that the two routes produced different products: the Russian overland tea acquired a reputation for arriving in better condition than sea-shipped tea, which spent months in a damp hold, and it was sold at a premium as caravan tea. The general point is the one this course has made repeatedly, that a trade route is not a thing that exists or does not exist but the current solution to a cost problem, and it persists exactly as long as it remains the cheapest way to move a particular good between a particular pair of places.
The real ending
If the caravans were still crossing Asia in the 1850s, what finished them? Three things in one generation, and they are all industrial.
The Suez Canal opened in 1869 and cut the sea distance from London to Bombay by around forty per cent, from something over 10,600 nautical miles round the Cape to about 6,300. Steamships, which no longer had to wait for a monsoon or beat against a wind, turned an annual cycle into a schedule. Between them, the canal and the steam engine did to maritime freight what the previous four centuries had only begun.
Railways did the same on land, and killed the caravan directly. The Trans-Caspian railway reached Samarkand in 1888, and the Trans-Siberian, begun in 1891 and through to the Pacific in the following decades, destroyed the Kiakhta tea trade by carrying the same cargo at a fraction of the cost and time.
Example. The Suez Canal cut the London to Bombay voyage from about 10,670 nautical miles round the Cape to about 6,330. If freight cost scaled with distance, what would that do, and why is the real effect larger?
The distance falls to of its previous value, a saving of about 41 per cent, so a purely distance-proportional freight cost would fall by roughly two fifths. The real effect was considerably larger, for reasons that compound. The canal route suits steamships, which can hold a course through the Red Sea where sailing vessels struggle, so it accelerated the switch from sail to steam on the eastern run and steam removed the monsoon timetable entirely. A shorter voyage means each ship completes more voyages a year, so the effective capacity of the fleet rises without a single new hull. Faster passage means less capital tied up in cargo afloat and less spoilage. And insurance and financing costs fall with time at sea. The general point matters more than the number: transport improvements do not simply reduce a freight bill, they raise the frequency and reliability of the service, and it is frequency and reliability that reorganise trade.
Now you. Given all this, what is the best short answer to "when did the Silk Road end?"
Answer
That it depends on what is meant, and that the question should be answered in parts rather than with a date. The period when overland Eurasian trade was the principal channel between China and the west arguably ended in the first millennium, since the sea was already carrying more by the ninth century. The period when a merchant could cross the whole span under one authority ended with the fragmentation of the Mongol khanates in the fourteenth century and did not return. The period when long-distance caravan trade remained commercially significant lasted until the second half of the nineteenth century, when railways and the Suez Canal undercut it. And local caravan traffic across parts of Central Asia continued into the twentieth century and in places has never entirely stopped. What did not happen at any point is a closure. If a single date is demanded, the 1860s to 1900s have far better claim than 1453, and the honest form of the answer is that a route is a cost calculation rather than an institution, so it fades rather than ends.
And the political geography closed. Russia took Tashkent in 1865, Samarkand in 1868, Khiva in 1873 and Merv in 1884; the Qing reconquered Xinjiang and made it a province in 1884. Central Asia stopped being a set of independent polities through which merchants passed and became the interior of two empires, then, in the twentieth century, the two sides of a sealed and militarised Soviet-Chinese border.
That is the ending: not 1453 but roughly 1869 to 1904, not a conquest but a set of engines, and not a closure but an obsolescence. It is less satisfying than the traditional story and it has the advantage of being supported by evidence.
One question remains, and it is the one the first lesson opened with. If the thing itself faded slowly and unremarkably in the age of steam, why does the name exist, and why is it more famous now than at any point in the past? The last lesson follows the term from Richthofen's map to a trillion-dollar infrastructure programme, and closes the account with a balance sheet of what actually moved and what it changed.