Wednesday, October 30, 2002

It is easy to think of knowledge as embodied in people – as part of human capital. Commonly we think that when you absorb new information, it becomes part of your knowledge. Our “knowledge” is “information” which we have in our minds. This is a nice simple formula, but neurophysiologists have spent a long time trying to figure out how the brain learns. It seems to be through changes in the neurons, and especially changes that result in different responses of neurons to synaptic inputs. Thus one might consider that human “knowledge” is information that has been represented in changes in the brain, where those changes allow us to recall and/or utilize the information.

We also think of books and libraries as repositories of knowledge. I think the implication is that the authors of the books have recorded the information that they know in the books. It is clear that I don’t “know” all the information that is contained in the libraries that I use. Do I in some sense “know” everything that is in the books that surround me in my office, where I can with relative ease locate that information?

I like the idea that I think with my brain and its “surround”. Thus, while my brain does not spell very well, my computer and I can produce this document with pretty good spelling, since information on how to spell words has been stored in my computer. It is of course stored in the form of physical changes to the discs used by the computer. This is an example of a more general concept – that we can embody information in machines, and indeed in objects. I would suggest that information so embodied also may be considered to be “knowledge”. Thus if a computer controlled machine is transferred to a developing nation, then the knowledge to accomplish the tasks performed by the machine exists in the nation, even if no person in the nation “knows” independently how to build the machine or carry out its tasks without the machine. Moreover, the embodiment of information in machines is not new and is not dependent on computers. Since the Jaquard Looms were invented, it has been clear that machines can be made to do things that in ancient times required human intelligence.

I would suggest in like manner, that information can be embodied in materials and such an embodiment can be considered to be knowledge. An example of this would be pharmaceuticals, such as diagnostics and therapeutics. I participated in a decade-long research effort to understand the etiology of acute respiratory disease in children. That effort was possible only because new reagents had become available that simplified the diagnostic process. I suggest that we can consider information on the diagnosis of respiratory disease to be embodied in these reagents, and that such embodied information can be considered “knowledge”.

Increasing people talk about “organizational learning”, suggesting that an organization can absorb information, converting it into knowledge. Clearly this can happen if the people or machines in the organization store the information. I would suggest however, that an organization can know something that none of its people knows, and that is not represented in its machines. Information that is encoded in changes in organizational structure and processes would seem to fit the concept of “organizational knowledge”, even though none of the members of the organization “knows” that information. (Incidentally, organizations can learn by adding personnel who know things, or be moving personnel around so that people with applicable knowledge are placed where it is useful.)

Formal organizations are one kind of institution among many. I think that we can consider that knowledge is embodied in other institutions such as markets, communities and associations. Thus where information is embodied in the people, machines, materials, organization or processes of these institutions, it may be considered “institutional knowledge”.

Thus I see the K4D challenge to modify the information that is embodied in people, machines and materials, organizations and institutions in order to improve the lives of people. This is a big challenge.

Monday, October 28, 2002

There has been a lot written about the nature of scientific and technological knowledge. By technological knowledge, I am here talking about the knowledge held by the modern technological professionals such as engineers, physicians, agronomists, etc. – not “traditional” knowledge about production processes such as is held by artisanal farmers and fisherman or traditional health practitioners. In summary S&T knowledge is that which relates to bodies of theory, that is obtained from controlled, replicated experiments, that is validated through professional peer review, and that is organized and made available to a professional community via its professional media (such as journals, conferences, and professional education).

Contrast this nature of S&T knowledge with, for example, the nature of bureaucratic knowledge characteristic of large formal organizations and their management information systems. Or with the nature of legislative knowledge, as developed by hearings, staff work, consultation with lobbyists and constituents, and the executive agencies of government. Or with the nature of judicial knowledge, as developed through the investigative powers of courts (in some systems) or of the advocacy-based trial system (in others), the accumulation of precedents, and the codification of systems of law. Or economic knowledge embodied in the prices developed in markets and other commercial processes. I could go on. I suggest that institutions generally have their own knowledge systems for the production, validation, organization and validation of knowledge.

These considerations are consistent with a view that knowledge is socially constructed, and I wonder if the most important way knowledge is constructed is institutional. Different institutions construe knowledge in different ways.

The K4D crowd seems often to assume that by focusing on scientific and technological knowledge they will most effectively promote economic development. It seems to me that this is at best a hypothesis in need of evidence. I suspect that development will depend on institution building including on knowledge systems in many institutions. S&T knowledge is important, and consequently so is building S&T institutions! But I would like to see K4D programs also emphasize institution building to improve bureaucratic knowledge (in formal organizations), economic and financial knowledge (in market institutions), legislative knowledge, legal knowledge, etc. The problem is to develop a process by which all this institution building can occur in parallel, with a reasonable allocation of resources among all the parallel efforts.
In the next few entries in this Blog I am going to do some riffs on things that have been bothering me on the topic of this Blog.

There is now a lot of interest in K4D, but it seems to me that a lot of it is based on economic analyses that are flawed. Some economists have used econometric methods with cross sectional data on the growth of capital, labor and GDP to conclude that much of economic progress is based on growth of productivity, and extrapolated that improved (technological knowledge) is the basis for improved productivity. I would agree that one can not improve the productivity of labor and capital without improving technology, but ascribing all the increase to that one factor seems to me unjustified.

It was once common to include “land” (or more generally natural resource endowments) among the determinants of GDP. It should be noted that in general natural resources are not “natural” – they are “stuff” that has been turned into resources by the accumulation of knowledge. Thus you have to know where something is, how to retrieve it, and how to use it, (and increasingly how to conserve it) to call it a resource. It seems to me that part of the increase in GDP has come from turning “stuff” into “resources”. This is not a new discovery; in North America the first economic applications of science and technology were in the discovery and mapping of resources – exploration, mineral prospecting, identification of soil, forest and fishery resources, etc. Without recognizing this reality, K4D programs will miss a key programmatic element needed to create wealth.

What do we do about institutions and policies? Most development professionals think that if you get the institutions and policies right, then development will follow. But institutional develop is not normally seen as a key element in K4D strategies, nor is improving the general policy making framework. In a future entry I will come back to the fact that knowledge is indeed embodied in institutions.

Perhaps because I worked in the Division of Research in Epidemiology and Communications Science of the World Health Organization, I am impressed by the need for problem identification as a key element in development. Epidemiological evidence allows health officials to define more cost effective strategies to improve health. Seems obvious, doesn’t it. Indeed, we can see HIV/AIDS is a grave global health problem because we have information on its incidence, prevalence, and lethality. Knowledge not only lets us define cost-effective strategies to reduce health problems, but allows us to identify those problems in the first place, and helps to set relative priorities among health problems. Similar arguments could be made for many fields. Again, I wonder how many K4D strategies devote sufficient attention to improving the knowledge base about societies’ problems.

While I am on the topic, let me point out that the focus on increasing GDP per capita is only a poor surrogate for solving societies’ priority problems. I think one of the key features of technological success is that they are associated with so rapid a decrease in cost of goods and services as to make econometric tools difficult to apply. Thus in the industrial revolution, the costs of key goods and transportation services decreased so fast as to make them almost disappear from consciousness as constraints. Similarly, we can see that there has been a real improvement in the lives of people in developing nations in terms of life expectancy, health status, food availability, access to information, etc. that seem to be invisible in the GDP per capita figures.

The basic point I am trying to make is that knowledge for development is a more complex field than improving technological knowledge to promote innovation and industrial productivity.

Sunday, October 27, 2002

I hear that the science and technology strategy for the World Bank is about to be approved and published.

The Bank has replaced its TechNet Network with a Knowledge for Development network or Community of Practice. The Community of Practice web site seems not yet to be operating, and links are still referred to the TechNet site, but it should be available in the future at:

I hear that Kerry Ann Jones is to be the new director of the Office of International Affairs at the National Academy of Sciences. She should be great at the job, bringing experience in the White House and USAID to the post.

Thursday, October 24, 2002

I am interested in reducing poverty in developing nations. I plan to use this Blog to write thoughts about that topic. I will especially focus on knowledge for development (K4D), and within that general subject science and technology (S&T) and information and communication technology for development (ICT4D).

This Blog supplements my web site:

It also relates to the Knowledge Economy and ICT for Development topics of the Development Gateway: