Friday, November 12, 2010

Transmitting technology


How do large technological advances cross cultural and civilizational boundaries? The puzzle is this: large technologies are not simply cool new devices, but rather complex systems of scientific knowledge, engineering traditions, production processes, and modes of technical communication. So transfer of technology is not simply a matter of conveying the approximate specifications of the device; it requires the creation of a research and development infrastructure that is largely analogous to the original process of invention and development. Inventors, scientists, universities, research centers, and skilled workers need to build a local understanding of the way the technology works and how to solve the difficult problems of material and technical implementation.

Take inertial guidance systems for missiles, described in fascinating detail by Donald MacKenzie in Inventing Accuracy: A Historical Sociology of Nuclear Missile Guidance. The process MacKenzie describes of discovery and development of inertial guidance was a highly complex and secretive one, with multiple areas of scientific and engineering research solving a series of difficult technical problems.

Now do a bit of counterfactual history and imagine that some country -- say, Burma -- had developed powerful rocket engines in the 1950s but did not have a workable guidance technology; and suppose the US and USSR had succeeded in keeping the development of inertial navigation systems and the underlying science secret. Finally, suppose that Burmese agents had managed to gain a superficial description of inertial navigation: "It is a self-contained device that tracks acceleration and therefore permits constant updating of current location; and it uses ultra-high precision gyroscopes." Would this be enough of a leak to permit rapid adoption of inertial navigation in the Burmese missile program? Probably not; the technical obstacles faced in the original development process would have to be solved again, and this means a long process of knowledge building and institution building.  For example, MacKenzie describes the knotty problem posed to this technology by the fact of slight variations in the earth's gravitational field over the surface of the globe; if uncorrected, these variations would be coded as acceleration by the instrument and would lead to significant navigational errors.  The solution to this problem involved creating a detailed mapping of the earth's gravitational field.

This is a hypothetical case. But Hsien-Chun Wang describes an equally fascinating but real case in a recent article in Technology & Culture, "Discovering Steam Power in China, 1840s-1860s" (link). There was essentially no knowledge of steam power in Chinese science in the mid-Qing (early nineteenth century). The First Opium War (1839-1842) provided a rude announcement of the technology, in the form of powerful steam-driven warships on the coast and rivers of eastern China. Chinese officials and military officers recognized the threat represented by Western military-industrial technology, but it was another 25 years before Chinese industry was in a position to build a steam-powered ship. So what were the obstacles standing in front of China's steam revolution?

Wang focuses on two key obstacles in mid-Qing industry and technology: the role of technical drawings as a medium for transmitting specifications for complex machines from designer to skilled workers; and the absence in nineteenth-century China of a machine tool technology.  Technical drawings were an essential medium of communication in the European industrial system, conveying precise specifications of parts and machines to the workers and tools who would fabricate them.  And machine tools (lathes, planes, stamping machines, cutting machines, etc.) provided the tools necessary to fabricate high-precision metal parts and components.  (Merritt Roe Smith describes aspects of both these stories in his account of the U.S. arms industry in the early nineteenth century; Harpers Ferry Armory and New Technology.)  According to Wang, the Chinese technical culture had developed models rather than drawings to convey how a machine works; and the intricate small machines that certainly were a part of Chinese technical culture depended on artisanal skill rather than precision tooling of interchangeable parts.

So communicating the technical details of a complex machine and creating the fabrication infrastructure needed to produce the machine were two important obstacles for rapid transfer of steam technology from Western Europe to Qing China.  But perhaps a more fundamental obstacle emerges as well: the fact that Chinese technical and scientific culture was as yet simply unready to "see" the way that steam power worked in the first place.  When steam warships arrived, acute Chinese observers saw smoke and fire, and they saw motion.  But they did not see "steam-driven traction", or the translation of kinetic energy into rotational work.  (This is evident also in the drawing of the treadmill water pump above; the maker of the drawing clearly did not perceive from the Italian drawing how the motion of the treadmill was translated into the vertical pumping action.)  Wang quotes a description from an observer in Guangzhou in 1828:
Early in the third month ... there suddenly came from Bengal a huo lunchuan [fire-wheel ship] .... The huo lunchuan has an empty copper cylinder inside to burn coal, with a machine on the top.  When the flame is up, the machine moves automatically. The wheels on both sides of the ship move automatically too. (37)
And another observer wrote in Zhejiang in 1840:
The ship's cabin stores a square furnace under the beam from which the wheels are hung. When the fire is burning in the furnace, the two wheels turn like a fast mill and the ship cruises as fast as if it is flying, regardless of the wind's direction. (37-38)
The give-away here is the word "automatically"; plainly these observers had not assimilated a causal process linking the production of heat (fire) to mechanical motion (the rotation of the paddle wheels).  Instead, the two circumstances are described as parallel rather than causal.

So the fundamental motive force of steam was not cognitively accessible at this point, even through direct observation.  By contrast, the marine utility of paddlewheel-driven warships was quickly assimilated. Chinese commanders adapted what they observed in the European naval forces (powerful paddlewheels that made sails unnecessary) to an existing technology (human- or ox-driven paddlewheels), and large "wheel-boats" saw action as early as 1842 on Suzhou Creek (40).

Wang notes that several Chinese inventors did in fact succeed in discerning the mechanism associated with steam power by the 1840s. Ding Gongchen succeeded in fabricating a model steam rail engine 61 centimeters long and a 134-centimeter model paddlewheel steamboat; so he clearly understood the basic mechanism at this point.  And Zheng Fuguang appears to have mastered the basic concept as well.  But here is Wang's summary:
Ding's efforts show that despite the circulating writings of a few experimenters, the steam engine remained a novelty, which was difficult to understand and probably impossible to reproduce. Interested parties were discussing it, however, but attempted to grasp it in terms of their indigenous expertise alone rather than more broadly understanding the new Western technology. (45)
In 1861, during the Taiping Rebellion, a senior military commander Zeng Guofan created an arsenal in Anqing for ammunition, and also set about to create the capacity to build steam-powered ships.  With the assistance of experts Xu Shou and Hua Hengfang, the arsenal produced a partially successful full-scale steamship by 1863, and in 1864 Hua and Xu succeeded in completing a 25-ton steamship, the Huanghu, that was capable of generating 11.5 kilometers per hour.  The Chinese-build steamship had arrived.

Here is how Wang summarizes this history of technology adaptation over a 25-year period of time:
The path from the treadmill paddlewheel boat to the Jiangnan arsenal's steamers was a long journey of discovery. Qing officials experimented with the knowledge and skills available to them, and it took time--and trial and error--for them to realize that steamboats were driven by steam, that machine tools were necessary to turn the principle of steam into a workable engine, and that drawings had to be made and read for the technology to be diffused. (53)
So perhaps the short answer to the question posed above about cross-civilizational technology transfer is this: "transfer" looks a lot more like "reinvention" than it does "imitation."  It was necessary for Chinese experimenters, officials, and military officers to create a new set of institutions and technical capacities before this apparently simple new technological idea could find its way into Chinese implementations on a large scale.

(The image at the top is one of the most interesting parts of Wang's very interesting paper; it establishes vividly the difficulty of transmitting technologies across different technical cultures.  The Italian drawing dates from 1607, and the Chinese copy dates from 1627.  As Wang points out, the Chinese version of the drawing is visually highly similar to the Italian original; it is a good copy.  And yet it fails to designate any of the technical features of how this treadmill-operated water pump works.  The pair of drawings are fascinating to examine in detail.)

No comments: