Thursday, July 15, 2010

The Nature of Technology

I have just started to reading Brian Arthur's The Nature of Technology: What it is and How it Evolves. I suspect the book will trigger a number of postings on this blog.

Brian Arthur is a very well known economist, most famous for his recognition that in some areas, especially including investment in technology, there can be increasing returns to scale. While this is an extension of his previous work, it benefits from his understanding of technology and I suspect from his own experience as an innovator.

Arthur understands "technology" as "means to accomplish purposes". Some of those purposes are quite narrow, as Privacy Informationthe technology of a hammer. Some are quite broad and diffuse as the technology for a super computer.

Technology as Embodied Knowledge

Arthur notes that there are both software and hardware aspects of technology -- devices and procedures for their use. He sees this as true of even a hammer, a very simple device and what appears to be a very simple procedure for its use. I would point out, however, that there is a very great difference between the way I use a hammer and the way a professional carpenter or cabinet maker would do so. Think about another similar technology, a golf club and think about the difference in the procedure for driving a golf ball used by Tiger Woods versus that used by the average duffer. The procedure for using a simple device well may require very precise procedures achieved through considerable analysis and requiring considerable skill.

I tend to like a definition of technology as "knowledge of how to accomplish a purpose". (Arthur recognizes that the word "technology" is used in several different meanings, as do I, and I also use a definition as "the organized body of knowledge about specific technologies.) Knowledge can be embodied in hardware devices, software, materials, people, or institutions. If you want to drive a nail and you don't have a carpenter's hammer, you can use another type of hammer, a rock or some other hard and heavy object. Surely good hammers make it easier to drive nails and thus can be seen as embodying some knowledge of how to drive nails easily and effectively. One can imagine a video instruction which could be used to teach a child to use a hammer well (or a video to teach a duffer to use a driver better), and that video might be seen as embodying knowledge to be later also embodied in the learner.

I would tend to include pure technique as "technology". Thus the technique used by a chiropractor to manipulate one's back to cure back ache or the technique used by a midwife to facilitate a difficult birth has very little of the "device" but considerable knowledge embodied in the skilled practitioner.

It is possible to have two technologies to achieve the same purpose but which embody different amounts of knowledge in their materials, devices and the human operators of the devices. Think of one cook preparing a dinner from frozen foods using a microwave versus another preparing a comparable dinner from raw foods using a gas stove, noting specifically how much less knowledge the cook needs using frozen food and a microwave.

Of course, things that embody technology can also embody other information. Obviously people have knowledge other than technological knowledge. A tool or a facility may well embody aesthetic information in its design as well as the purely technological.

I would say that a device may incorporate ethical values. Thus during World War II, American fighter planes were more heavily armored than Japanese fighters. The heavier planes were less maneuverable but safer for the pilots. The designs seem likely to reflect not only accidental differences and differences imposed by the different economic strengths of the two countries and strains on their economies imposed by the war, but also differences in the value of the lives of the pilots and on the maneuverability of the weapons provided to those pilots.

Technology and Social Institutions

Steve Lansing has studied the system for the distribution of irrigation water in Bali. One of the phenomena he noted was that when "modern" irrigation managers from Java took over the management of the system from the traditional managers, the priests of the water temples, the pest problems in the irrigated rice fields increased dramatically and yields of rice dropped. None of the farmers and none of the temple priests could explain why their traditional system moderated the growth of insect populations, but it did. I would say that the technological knowledge of how to reduce insect populations and increase rice yields in Bali was embodied in the traditional ritual processes that had evolved over generations for the construction of new elements of the irrigation system and for the allocation of irrigation waters. This is true even though no individual could articulate the reason the processes worked or even give a complete description of the processes.

Arthur goes further, identifying some institutions as technologies in that they are means to accomplish purposes. Thus a college may be seen as a means for accomplishing the education of its students. While a college may have a physical aspect -- buildings, libraries, etc. -- a college without walls might equally be seen as accomplishing the educational purpose. Arthur recognizes that we usually require a device to be involved to invoke the term "technology", but holds that it may be useful to see some social institutions as technologies. (I wonder if some institutions such as the family or the small group which are rooted in human evolution and not cultural innovations would qualify.)

As a graduate of UCI's Graduate School of Administration, I rather like the idea that a formal organization such as a business can be seen as a technology. That view makes clear that organizations can be engineered, using knowledge accumulated from the study of organizations and organizational behavior, to accomplish purposes more efficiently and effectively.

Technological Systems and Recursion

Arthur points out that the word technology can be used to describe an airplane, but also to describe each of the components of the airplane, or to describe an aircraft carrier with all of its component parts as well as the fleet of airplanes it carries. He makes an important and subtle point that building technologies as combinations of other simpler technologies is both pervasive and extremely useful.

He suggests that eventually the reduction of a technology gets to elemental technologies. The hammer and nail example would be seen as composed of a nail and a hammer, with the hammer perhaps composed of a handle and a head. I would note however that a nail or a hammer head is the product of another technological system, indeed there are factories devoted to the production of nails or hammer heads. Thus the elemental technologies may not be so elemental. The hammer can be seen to embody a huge body of knowledge relating not only to the best shape and weight for a hammer, but the best materials from which to make a hammer and the manufacturing technology necessary to prepare those materials and combine them.

Other Kinds of Technological Knowledge

Changing examples, consider the antibiotic. The utility of an antibiotic depends on using it to attack bacteria and not viruses, and indeed to attack a specific bacterium or one of a class of bacteria known to be sensitive to that specific antibiotic. Thus the physician needs to embody "technological" knowledge to determine the specific purpose for which to prescribe a remedy.

Epidemiological research, similarly, determines the public health problems which public health officials can address with specific technological remedies. Thus a recognition of the areas that may become endemic for malaria with climate change will lead public health officials to choose a variety of techniques for vector control, case finding, health education, and treatment of infected, each with its own devices and procedures.

Arthur points out that technology is used to transform inputs to outputs. He uses the example of the radio which takes as its input an electromagnetic signal transmitted through the atmosphere and produces sound recreating the original program.

If one returns to the example of the manufacture of a hammer, the input might be wood for the handle and the metal for the head. I note however that not all woods are equally good for efficiently making good handles nor all minerals equally good for making the metal used in a hammer head. There is knowledge involved in "raw materials", including the knowledge of what the material can be used for, where it is to be found, and how it is to be obtained efficiently. In some sense this too is technological knowledge.

Why technological innovation shows positive returns to scale

Brian Arthur is credited with making the economics profession focus not only on situations with decreasing returns to scale, but also to increasing returns. He does so again in the context of this book. Arthur suggests that new technologies are generally assemblies of existing technologies. Thus the more technology exists in a society, the more pieces there are with which to build new technologies. I can think of examples in which that is true. Having silicon factories which can manufacture chips together with computers and software to help design chips and expert chip designers means that it is relatively easy to develop an new electronic device that is built around chips, as has happened with many innovations in consumer electronics.

One of the conundrums of international development is the difference between the rapid reconstruction of Europe and Japan after World War II had decimated its manufacturing plant versus the new nations of Africa through decolonization which had little manufacturing capacity then and still have little capacity today.

I suggest that a while both societies were lacking in devices in the past, there was a great deal of technological knowledge in the European population 65 years ago and very little technological knowledge in the populations of Africa. Of course there was a similar difference in the knowledge of how to construct and operate "modern" institutions in the two continents. Still if you know how to put together modern technological systems you ought to have an advantage building or rebuilding a productive economy.

Thinking About Invention

Either Arthur does not treat invention well, or I have not gotten to that part of the book yet. It seems to me that while many technological innovations are simply incremental improvements ot existing technologies but a few are game changers. I think of the radical innovations as those with the possibility of changing a technological paradigm.

Think about the effort to create an HIV vaccine. Arthur stresses that technologies are built around properties. HIV vaccine technology, if one can be developed, will be built around a material with the property that it safely induces HIV immunity in an immunized person. Since some few people appear to not become infected with HIV in spite of repeated exposure, it seems likely that there must exist one or more substances with the property of inducing immunity, snf therefore other such substances could be developed. Unfortunately, no one knows the identity of such a substance. A number of people have been working hard for decades to find such a substance. When an appropriate substance is identified and proven both safe and efficacious other technologies can and probably will be applied to the production of large amounts of the vaccine and mass immunization campaigns.

The point is that the invention of a vaccine against HIV/AIDS is not like the development of a new model automobile or a new model television, even the move from analog to digital television. Rather it involves the discovery of non-obvious properties that must be found for the technology to exist, properties which indeed may not be achievable.

Increasingly such inventions arise from basic research for the discovery of new properties.

No comments: