A PHILOSOPHY OF COMPUTING FOR DEVELOPMENT

Back

The following was written as part of the course ICT for Development held at Vrije Universiteit. Date: May 2018. Author: Hans-Dieter Hiep


In this essay, I will argue how philosophical viewpoints of computation compare to notions of development. By computation I mean the process of step-by-step calculation. By development I mean what most readers understand by saying—development in the Third World or in Developing Countries or in the Global South. We shall first reflect on philosophy of computation, and we shall derive basic consequences. We shall then reflect on Information and Communication Technology for Development (ICT4D), and relate our philosophy to these concepts.

The reader is warned to proceed reading this essay with care. I am not a philosopher nor a social scientist. My background is in, or I feel myself affiliated with, theoretical computer science (TCS). TCS embraces discrete mathematics: formal languages, term rewriting systems, arithmetic, combinatorics, abstract algebra, higher-order logic, proof theory, and other methodologies.

We repeat: "computation is the process of step-by-step calculation." Many models of computation have been presented before. For example, Turing machines perform step-by-step calculations by moving between internal states and at the same time acting on some external tape of symbols. Term rewrite systems model computation by pattern matching in terms, rewriting them according to a fixed set of rules applied within context, resulting in sequences of terms. More esoteric models of computation include cellular automata, among many others.

Samson Abramsky raises valid questions concerning computation (Abramsky, 2014). Why do we compute? This question might seem silly: we compute to gain more information. This answer has two problems. First, a computation cannot give one "more information," since that is a contradiction in Shannon's information theory—the entropy of an isolated system never decreases! The second question is: if a computation always gives the desired result, why does one need to perform the computation in the first place? These two questions are dissolved by a shift in perspective.

One should think of open subsystems, instead of isolated systems, between which a flow of information is possible. A subsystem should always be regarded as part of a larger context: whatever surrounds a subsystem is called the environment. Such a subsystem can increase in information by consuming energy from its environment—a property which is well-known (Landauer, 1961). One increases information by deleting the irrelevant. In the words of Abramsky, "the actual usefulness of computation lies in getting rid of the haystack, leaving only the needle."

Moreover, Abramsky describes the differences between 'explicit' and 'implicit' knowledge. A computation is a process that converts implicit knowledge, say as encoded in the rules of operation, into explicit knowledge, encoded as its result. The cost of such a conversion process is measured by the consumption of resources—the number of steps and the required work space it takes. Clearly, lots of irrelevant information is deleted as the workspace is reused over time.

Bertrand Russell has left an impressive mark on modern mathematics with Principia Mathematica. Russell was an essayist, and among his works is the essay "In Praise of Idleness," (Russell, 1932). The main point of that essay is that "I want to say, in all seriousness, that a great deal of harm is being done in the modern world by belief in the virtuousness of work, and that the road to happiness and prosperity lies in an organized diminution of work." This is in stark contrast to, what is believed by many, that hard work is a virtue. The reason for that thinking, according to Russell, is that hard work used to be necessary to for ones own subsistence. However, "modern technique has made it possible to diminish enormously the amount of labor required to secure the necessaries of life for everyone." He demonstrates this point by modern wartime, in which only a small portion of the population is actually productive, and all others waste their energy on efforts related to war.

But what is work, actually? There are two classes of work according to Russell. The first class of work is the manipulation of physical matter near the earth's surface. The second class is the act of telling other people what work they should do. Let us call the first class direct work, the second class indirect work. From the perspective of anti-foundational mathematics (Aczel, 1988), our definition of work seems to include the possibility to have a non-well-founded sequence of indirect work. For example, a circle of people where each person tells the next what to do, possibly with some branches off to people that do the direct work. Many social constructions of today still resemblance such cyclic structures, as witnessed by the bureaucracies in government and industry. Russell mentions that these two classes of work are directly reflected in social groups: those who do direct, physical, blue-collar work; and those who do indirect, advertisment, white-collar work. There is still a minor third social group: those who do nothing, but extract from the other two groups by letting them work.

From history of computing in the West, we know that classical digital computers were employed to increasingly au-tomate many business and government activities over time. Indeed, one of the first major application of computing machinery was to automate the US population census. These efforts have alleviated many people from doing direct work, in the case of human calculation, and indirect work, in the case of the management of such enormous tasks. Indeed, for many organizations, automation and digitization projects are still beneficial precisely by the reduction in work and the corresponding decrease of labor costs.

Idleness is an important concept in development in the Third World. The concept of the Third World is strongly tight to the geopolitical situation during the Cold War, in which president Truman of the USA identified three worlds: the First World, being the United States of America and its allies; the Second World, being the communist dangers such as the Soviet Union and China; and the Third World. President Truman convinced decolonized countries to recognize themselves as 'underdeveloped' and allow long-term assistance from the USA (Potter, 2017). Indeed, "the American advisers to the Thai government of the 1950s [tried] to prevent the monks from preaching the virtues of contentedness, which was seen as retarding modernisation." Contentedness is necessary for idleness, since one can only be idle if one is content.

Clearly, we see how the notion of 'underdeveloped' emerges from the Western perspective of hard work. In the 1950s, 60s and 70s, as western Europe was recovering from the past two World Wars by pursuing hard work, the vision emerged to replicate this behavior from the North to the South. As Potter writes: "development became synonymous with economic growth. Growth theory and models of growth ... were developed and pursued ... in an effort to achieve 'growth'."

However, no concern was given to the distribution of such 'growth,' with the possibility of increasing social unrest as a result. Between 1970 and 2000, the mission sophisticated to include better distribution and virtues of 'freedom' up to where it developed into 'sustainable development,' where development meets the needs but does so without negatively affecting future generations. These movements have lead to several criticisms, among which the arrogant Eurocentrism is the most important.

Exemplary of Eurocentrism is the real economic result of development: according to Potter, each year the continent Africa imports US $132 billion in investment and loans, but exports US $192 billion in dividend and repayments. If we reflect this criticism of development on Abramsky's idea, we can see a different image of what is happening: while development aims to increase the freedoms and material quality of living (cf. increase in 'information') for Developing Countries, it does so by consuming 'energy' from its environment—and in economical terms, such 'energy' consumption comes at an economical cost: dividends and debt repayments. In terms of Russell's social groups, the North acts as an idle group, extracting the work of the South by letting them develop. However, a more promising side-effect of development is that it deletes the irrelevant: work. Thus, 'the poor' are able to work less, and prosper in culture, science and social life.

A particular field of development is the recently emerged ICT4D, Information and Communication Technology for Development. It simultaneously aims for development and research by applying techniques of ICT systems, well-known and heavily applied in the North, and by reflecting on the use of ICT in Developing Countries. According to (Walsham, 2012) the field has an ethical goal of using information technology to support 'the poor' and not the economically well-off. In addition, the field has a critical agenda of finding out who benefits from technology and who does not. Walsham writes, "[f]or example, the Internet is currently a rich person's tool. Can the spread of smartphones, for example, enable the poor to access the Internet and how can this help to reduce poverty?"

One may support this agenda, even though the mission of improving the economical situation of 'the poor' falls under the guise of Eurocentric arrogance. Once 'the poor' in Developing Countries depend on ICT, the West that has introduced technology may now exert a position of power. This is not necessarily bad or unethical, since, for example, it allows for idleness: anyone with more time free eventually uses that time to develop personally, thereby allowing for a replication of the Enlightenment of the West. I would even go further, and strive to educate and enlighten 'the poor' into developing ICT systems themselves. This grants them a position of independence with respect to technology, and allows them to extract profits gained from information technology directly without any interference from the West. To do so, an ICT system should be free and open for modification, and engage the user into learning how the system works.

A known problem within ICT4D is that "the [ICT4D] research community is not unified on how to harmonize the diffi-cult and sometimes competing goals of conducting experiments, producing social change, and studying the phenomena of ICT use in developing countries" (Heeks, 2017; Dobson, 2013). On a basic level, we suggest that ICT4D researchers maybenefit from appreciating our philosophy of computation as it relates to reduction of work and improving the idleness of 'the poor'. A benefit of studying ICTs in Developing Countries over the West is their lack of highly involved and circular social structures that involves indirect work, as we experience with legacy bureaucracies in the West. Hence, applications of ICTs reduce mostly direct work. Additionally, an ICT4D project should answer the question: why do we compute? This question can be answered according to the ethical goal of ICT4D: we compute to obtain information, but we must do so fairly and evenly distributed. One should be critical of computation: who benefits from obtaining information, and who does not?