February 2015

No, not the drugs forum, the real one.  I just read a book about it, Silk Road: A new history, by Valerie Hansen.


There’s been a lot of archeological research in the last fifty years, and it overturns much of the received wisdom about the Silk Road. E.g., let’s take the summary from Wikipedia:

Trade on the Silk Road was a significant factor in the development of the civilizations of China, the Indian subcontinent, Persia, Europe, and Arabia, opening long-distance, political and economic interactions between the civilizations. Though silk was certainly the major trade item from China, many other goods were traded, and religions, syncretic philosophies, and various technologies, as well as diseases, also travelled along the Silk Routes. […]

The main traders during antiquity were the Chinese, Persians, Greeks, Syrians, Romans, Armenians, Indians, and Bactrians, and from the 5th to the 8th century the Sogdians. During the coming of age of Islam, Arab traders became prominent.

The major edit: strike the focus on “trade”. Thanks to the dry conditions in the Taklamakan desert, we have extensive written records from certain cities and periods, and trade was– if you’re picturing huge, constant caravans– underwhelming. The size of a party was generally under 10 persons. Only a third of the parties had horses. Groups arrived and departed less than once a week. (We have these records because trade was taxed and regulated. To move within China you had to have a document listing the people and animals in your group and the cities you intended to visit.) It was also incredibly slow: if you made 20 miles a day you were doing pretty well.

The biggest actor in the height of the Silk Road (the first millennium CE) was the Chinese government. It was the major source of product (silk bundles used both as cloth and as currency), of coins, and of demand (what it mostly wanted was horses). When the Tang lost control of the Taklamakan, trade declined greatly.

Also strike “Romans”. The silk route was essentially a line of contact between China and Persia, with an important spur down to India. Quite a few Persian coins have been found in China; no Roman ones. (By contrast Roman coins are easily found in India.) However, the Byzantines got in on the action. The Romans had silk, but most of it came from India or the island of Cos.

And don’t picture people traipsing from Antioch to Xīān. Most trade was local– a merchant would make a circuit of 500 miles or so.

Once you accept the scale and nature of the trade, I’m afraid Hansen ends up undermining her own point. She emphasizes over and over that “merchants” are rarely mentioned in official documents, that full-time traders were small beans, that pilgrims, Chinese garrisons, and official envoys were more important economic factors. Well, yes, that means that goods were traded. The people doing the trading might not have called themselves traders, but large amounts of materials were entering and leaving the region.

It’s well documented (and not neglected in the book) that religions traveled along the silk road– especially Buddhism, but also Zoroastrianism (the old Persian religion), Islam, Judaism, and Nestorian Christianity. And again, by her own account, there was a Central Asian community in Xīān, and a Chinese stereotype of Central Asians as rich merchants. These point to a fairly intense level of contact– China was influenced by people and ideas coming from Central Asia far more than those coming from, say, India. (E.g. the Indic writing systems and Theravada Buddhism reached Southeast Asia but not China.)

Plus, don’t underestimate local-scale trade. There are Shāng dynasty tombs from about -1000 containing Central Asian jade. This was long before there was any military presence in the Taklamakan; but jade was increasingly valuable the farther you got from its origins, so it could migrate east in small steps. The major long-distance trade items on the silk road were things like spices, silk, paper, jade, and sal ammoniac– items that were lucrative even in small volume. Thanks to the magic of geographic pricing, the silk road could move items a long way without any individual travelling its whole extent.

Finally, the total tonnage transported might have been low, and mostly inessential luxury goods, but, well, what else would you expect? Large-scale trade in staples like grain existed, but within empires (Egypt to Rome, south to north China). Luxuries are big motivators– the European rush to reach the East was driven by a taste for spices and other Asian goods.

In UI design, a user model is how the user thinks something operates. This is my user model for the shower in my apartment.


  • I turn the faucets to a desired temperature.
  • The settings are transferred by mechanical means to the basement.
  • Sensors there determine the position of the controls. These are converted into verbal instructions and communicated by telephone to an old man, who actually adjusts the amount of cold and warm water.
  • The old man is deaf, grumpy, and frequently away or drunk, we don’t know.
  • He’s of mixed Italian and Serbo-Croat descent, and sometimes he hears “cold” as caldo (Italian ‘hot’), or “hot” as Serbo-Croat hladan ‘cold’.
  • Sometimes salesmen call; he takes whatever they say as more commands and changes the water setting. Or maybe it’s just sheer cussedness.
  • There is a range of settings that produces a warm, comfortable shower. These settings are marked “DO NOT USE, FIRING OFFENSE” and he strictly avoids them. However, he can sometimes be coaxed into this range by confusing him with frequent, contradictory orders.

This user model explains and predicts the shower behavior quite reliably.

For my book, I just read Robert Temple’s The Genius of China (1998), which is a popularization of Joseph Needham’s decades-long project to recover and inventory the science and technology of China.

A drum roll for... the dromon

A drum roll for… the dromon

Now, I’m 90% satisfied with the book. For most of its history, China was not only ahead of Europe– which wasn’t that hard to do– but ahead of the Mediterranean and Middle East. It had more intensive agriculture, more sophisticated government, a more inclusive religious atmosphere, and a jump start on any number of technologies: paper, porcelain, gunpowder, printing, the wheelbarrow, the spinning wheel, canal locks, watertight bulkheads, cheap cast iron, the compass, even mathematics and equal temperament in music.

Nor is this, as a general principle, surprising: you’d expect the planet’s richest and most populous civilization to be its most advanced, and so it was. It’s not a bad bet that in a few centuries, when we finally enter galactic civilization, we won’t be known as Terrans but as Dìqiúrén.

But sometimes I want to ask… did anyone check over Needham’s work? To support the idea that the Chinese had man-lifting kites, Needham cites the alchemist Gě Hóng, writing in the 4C:

The Master said, Some have made flying cars with wood from the inner part of the jujube tree, using ox leather straps fastened to returning blades so as to set the machine in motion. Others have had the idea of making five snakes, six dragons, and three oxen to meet the “hard wind” and ride on it, not stopping until they have risen to a height of forty lǐ [13 miles]… This account comes from the adepts and is handed down to ordinary people, but they are not likely to understand it.

Needham concedes that we have no evidence of the sort of machines Gě describes, but as for the kites, he thinks “there would have been really nothing to prevent it.” Reeeeally, Joe? That’s the sort of dodge you use when you really really want to believe something. Temple helpfully provides a picture of a hang glider to help the claim sink in.

Surely flight is one of the things attributed to sages all over the world? The Siberian shamans and Castaneda’s Don Juan were supposed to be able to flit about using the spirit world; TM dudes are said to levitate; witches fly on brooms; the Monkey King could jump 108,000 lǐ at a time. In Matthew 3, the devil takes Jesus to a mountain tall enough to see “all the kingdoms of the world”. Gě Hóng goes on to support his notion of a high gale by appealing to how dragons fly. Reading an alchemist literally is generally not a good idea.

In other cases my question is, if you pored over the Western and Arabic historical record, mightn’t you find a large number of surprises? The Greeks and Romans were loquacious on their philosophy and politics, tight-lipped on their engineering. E.g. knowledge of the Antikythera gear work was lost for millennia.

A few pages after the hang gliders, Temple remarks “For most of history, Europeans used ships which were drastically inferior to Chinese ships in every respect imaginable. They had no rudders, no leeboards, no watertight compartments, single masts, and square sails…” Which is simply wrong. Roman ships had rudders and multiple masts. The Byzantine dromon, shown above and used throughout medieval times, had multiple masts and triangular sails.

Needham often suggests that any particular Chinese discovery must have found its way to the West. This is frustrating, as in fact science is full of multiple independent discoveries. Now, sometimes he has good clear evidence: e.g. he has a 1787 quote from Ben Franklin suggesting that mail ships be built with watertight bulkheads “after the Chinese manner”. But the monogenesis of ideas is a terrible heuristic.

Temple’s cheerleading gets tiresome at times— e.g. the last section, on warfare, says “No nation in the world could match the Chinese expertise in warfare for two millennia.” Except, you know, China was conquered twice by northern nomads, and northern China several times more. And China, for all its might, was remarkably unsuccessful at expanding into Korea and Vietnam, despite multiple tries. I think the problem here is that Temple confuses invention with mass deployment: they had land mines! burning gas! cannons! repeating catapults! Well, yes, but it’s far from clear that these were manufactured in sufficient quantity, or used with sufficient skill. Everything in China was at huge scale, but so were the logistical problems. So far as I can see, China’s government was always underfunded, and its army was subpar for an empire its size (not least because its elite was– unusually for the premodern world– not drawn from the military).

(And yes, I’m aware that for some of those conquests, the nomads used Chinese troops and engineers.  On the other hand, the Sòng, back when they controlled all of China, were completely unable to get the Sixteen Prefectures back from the Khitans, which doesn’t say much for Temple’s claim of unbeatability.)

Anyway, it’s still a fascinating book.  As I say, I don’t have problems with 90% of it.  It’s just that a few bits rang my skepticism meter.