Review of an African Rural Internet Network and related Academic Interventions

Nelson Mandela Metropolitan University, Port Elizabeth, South Africa

INTRODUCTION

LinkNet operates a rural internet network in Macha, Zambia, Africa. Macha is a small and resource-limited rural village. LinkNet is based upon a holistic and respectful vision aimed at developing the potential of the local community, which plays a leading role in the progression of the local internet network. The communications network is established within the setting of a co-operative not-for-profit institute.

Setting up internet access and connecting rural communities in Africa is an involved and complex endeavor. Not only are basic communications and energy technologies lacking, but also a diverse array of constraints have to be conquered. The integration of academic research findings is a challenge. This article addresses various aspects of these challenges. Most specifically it deals with an array of academic resource issues, like the availability of guiding literature in the fields of context, culture, orality, information and communication technologies (ICT) and its engineering, the digital divide and digital exclusion, applications, costs, and management practices. Lastly the article introduces local perspectives on the value of academic interventions and concludes with the need for conceptual frameworks incorporating all of these complex aspects in a culturally adaptive way.

BACKGROUND

Rural African communities are globally typified as lacking in development and in need of change (Heinemann, Prato & Shepherd, 2011). There is wide international support for the need for development of the least developed regions on the earth, for diverse reasons (United Nations, 2009).

Traditionally, engines of change appear in the form of development aid or missionary work (Meganck, 2010). Development aid disperses its funding through projects, often bringing together experts in capacity building for a specific purpose. Development projects are part of thematic development programs, and they involve distinct phases like planning, grant writing, execution, reporting and assessment. During the assessment phase, measurements and evaluation matches outcomes with goals set for the project.

Development projects are operational in a wide range of disciplines. Most operate in a vertical manner and link to international priorities like the Millennium Development Goals (MDG). Goals set by funding agencies mostly define the development agenda. Priorities of funding agencies are set by international politics or individuals. Funding agencies are mostly bilateral or multilateral institutions, or private, philanthropy foundations. Beside institutionalized development projects, remittance of funding from the diaspora abroad fuels separate development activities in developing countries (Kozul-Wright et al., 2010). This flow grows, with further growth expected through funding streams in social venturing.

Literature shows that equal partnerships and inclusiveness in development of local interventions is a key requirement for long-term sustainability (InfoDev, 2008). The need for incorporating human values and culture is being recognized (Miller & Larson, 2005), and there is openness to exploring development using a socio-technical approach (Amadei & Wallace, 2009; Kam, 2012). Although successes have been reported for development projects, many reports depict failure. Most development projects do not reach their set goals, not even closely. Resentment is growing toward traditional approaches to development: there are many critics with as many diverse positions, e.g. Khoza (2005), Moyo (2009) or Adamson (2012). The challenge remains to not only find good outcomes as defined by outsiders but also to find good outcomes in the context of local authenticity, limited resources, and community priorities.

The use of information and communications technologies (ICT) in communities or by individuals at the "Bottom of the Pyramid" is the subject of much and diverse study. Literature focuses on business and empowerment and on ICT in general and information systems as the most common technology objects of analysis, with a growing trend toward mobile phones (Gomez, Baron & Fiore-Silfvast, 2012). (The term 'Bottom of the Pyramid' - or BOP - has been put into the mainstream by Prahalad; for instance in his book The Fortune at the Bottom of the Pyramid, Prahalad (2009) refers to the billions of people living on less than $2 per day, also popularly dubbed the other three billion .)

The global impact of ICT is enormous. However initiating, implementing, operating, and scaling up ICT use in rural sub-Saharan Africa seem to be routed through a particular minefield of challenges. Apart from quantitative engineering aspects, a multitude of qualitative constraints have considerable effect (Stam, 2011) involving environmental, skills, and cultural ingredients. Issues such as high costs of inputs and the means of access control contribute to the array of challenges.

Top-down, technology-centric, goal-diffuse approaches show unsatisfactory development results (Dodson, Sterling & Bennett, 2012). A mix of reasons is offered as hampering the local adoption of technology. Not many projects show the capacity for sustainability nor the properties of scaling up. With the exception of anthropological approaches, accademic interventions researching how qualitative aspects and engineering interact are few and far between. There is little evidence on the modalities of long term local adoption and respectful integration of access projects in rural areas in Africa, or how they interact with science. This lack of evidence hampers understanding.

METHODOLOGY

From the initial set-up of the internet network in 2004, LinkNet network grew to a considerable size, involving well over 100 wireless units. From 2004 to 2011, the network was linked to the public internet through satellite connections. In 2011, the network was terrestrially linked to an Internet Service Provider in the capital city, Lusaka. The engineering details and use of the network has been studied in detail through interaction with academic researchers but little research has been conducted on the qualitative aspects of the network.

Our approach is to conduct a broad and qualitative literature review to shed light on the interactions at the internet network initiative at LinkNet/Macha Works in rural Zambia and to corroborate that with academic literature. The literature review not only considers ICT(4D) literature but also casts its net much wider, exploring many aspects that arise during the introduction of ICT in rural areas. The outcomes of this literature review are then contrasted with longitudinal ethnographic research findings in the rural community of Macha, in Zambia's Southern Province.

RESEARCH

Local Culture

Macha is a community in rural Zambia. It is a typical rural community, being a considerable distance from any town or economic center. Macha resides in communal lands. Also in the community are medical and educational institutions based upon church-administrated title-deeded land. The pastoral lifestyle of residents centers on people-interactions, subsistence farming, and livestock.

A string of multi-disciplinary papers involving quantitative and qualitative reports corroborate evidence of longitudinal research depicting the progress of this rural community since 2004. For more than 20 years, high quality research has been ongoing at Macha, including ground-breaking publications focused on results from academic interventions by the Macha Research Trust (Thuma, 2010). The range of publications spans multiple studies in measles, malaria, tuberculosis, and HIV, often in collaboration with universities from the USA (Moss et al., 2011) and Europe. Although these medical studies recognize the importance of the ICT networks in Macha, only one study specifically reports on ICT, dealing with the use of GSM to send weekly information of rapid malaria tests used and number of positive diagnoses using SMS (Kamanga, Moono, Stresman, Mharakurwa, & Shiff, 2010). Personal interactions show the internet as a life-line in the medical research community in Macha, however, none of the medical research publications quantifies or qualifies the specific benefits of the internet network.

With regard to a literature base, a few (small) books have appeared through the years, often written by missionaries or doctors positioned at the Macha Mission. These books are difficult to find. In 2011, an e-book appeared from this rural area (Stam, 2011). On a provincial scale, books on Tonga culture do exist though hard to find. Elizabeth Colson, an American anthropologist, published extensively on the Tonga, e.g. Tonga Religious Life in the Twentieth Century (Colson, 2006). She compiled hundreds of published and unpublished references on the Tonga-speaking people in Zambia and Zimbabwe (Colson, 2008). Another source of history is Hobson's Tales from Zambia (Hobson, 1996), and Gewald et al.'s One Zambia, Many Histories (Gewald, Hinfelaar, & Macola, 2008). However, these publications are relevant for contextual analysis only, none mentioning ICT.

A limited amount of literature is aimed at directly benefiting local practitioners in the rural African context. The non-discursive expressions of scientific knowledge, reduced to abstractions in English texts, seem to have little discernible effect on - even prohibiting the inclusion of - oral societies (Stam, 2013). The almost negligible African representation in formal academic publications in almost any field creates a defacto dependence on out-of-context, foreign scholarly direction (Gitau, Plantinga, & Diga, 2010). Recognizing this divide, and that texts are necessarily directed to western consumption, David Maranz, in his book ' African Friends and Money Matters ' declares 'the hope that [the book] will contribute to Westerners having greater respect for a unique economic system that accomplishes its main purposes very well' (Maranz, 2001). In literature, a number of books give a whole sweep of history and evolution of thought. Their generalizations can be helpful in assessing long term dynamics of change. For instance, Morris observes 'It is one thing .. to sit around tinkering, but it is another altogether for .. ideas to catch on and change society. That, it seems, requires some sort of catalyst' (Morris, 2010).

Conducting academic interventions that address technology from within in order to benefit rural Africa is a formidable challenge. Culture is the context in which things happen; out-of-context, technology lacks significance (Trompenaars & Hampden-Turner, 2011). Unfortunately, the non-discursive requirements in academic publishing hamper Third World scholars from producing, and publishing, scholarly publications, or even accessing them. Canagarajah writes that literature reviews thrive on conventional wisdom, incorporate overwhelmingly North American and Western European thinking and its intellectual hegemony, and embed international power relations (Canagarajah, 1996). Dourish & Mainwaring (2012) link computing into a colonial intellectual tradition and identify the problems that arise in consequence.

Dilemmas

Cross-cultural knowledge is hard to come by and the consequences often baffle people involved, leading to hostilities as cultural identities are defended (Lanier, 2000). Dilemmas posed due to cultural diversity - in the context of literature and science - have been noted by various authors. Trompenaars & Hampden-Turner (2011) observes that although integrated technologies have a logic of their own that are not affected by culture, the objective dimension can have totally different meanings to different local human cultures. In the same book, the authors note that seminal literature's depiction of 'one best way to manage and organize' is based upon a believe system that such universalism is scientific, when in fact such position is a cultural preference. Not surprisingly, literature tends to depict a depressing state of (rural) Africa, which does not necessarily correlate with local perceptions. South African academic Khoza, in his book Let Africa Lead , writes, 'Those of us accustomed to mixing with outsiders are used to hearing a few polite and tentative remarks about the "problems of Africa after independence", followed by an embarrassed silence. Aid-givers celebrating their selfless assistance to poor old Africa are wont to lay misgovernment and corruption at our feet, like a corpse at a wedding feast' (Khoza, 2005).

Ubuntu Culture

Sub-Sahara African culture is based on Ubuntu. Khoza explains the culture as an epistemology and humanistic philosophy, a metaphor embodying the significance of group solidarity (Khoza, 2005). It is key to all African values, involving collective person-hood and collective morality. Tutu writes. '[Ubuntu] also means my humanity is caught up, is inextricably bound up, in theirs. We belong in a bundle of life' (Tutu, 1999). He contrasts western philosophy and Ubuntu through 'it is not "I think therefore I am". It says rather: "I am human because I belong"'.

Others take definitions from radical humanism paradigms. Mbiqi and Maree define Ubuntu as the sense of solidarity or brotherhood which arises among people within marginalized or disadvantaged groups (Mbiqi & Maree, 2005). Louw (2002) regards Ubuntu as a response to multiculturalism, with specific reference to South Africa. He defines Ubuntu as 'an African or African-inspired version of an effective decolonising assessment of the other.' The claim in these assessments transcends absolutism without resorting to relativism and involves respect for religiosity, agreement on criteria, and the necessity of dialogue of beliefs. Colson (2006) notes that especially the role of beliefs is undervalued, and in her work she gives insight into the sheer complexity of the arena of religiosity. Despite it being the cultural expression of hundreds of millions of people, literature on Ubuntu is often regarded idiosyncratic. Practical implications for academic interventions, organizations, and change theories, have been explored in literature, albeit sparsely, e.g. Van der Colff (2003) and Van den Heuvel (2008). On Ubuntu and African Renaissance a small but steady stream of exploratory and critical books appear, like Moeletsi Mbeki's Advocates for Change (Mbeki, 2011), however, with few bookshops in most of Africa, they are difficult to access. With the expansion of the internet in Africa to the blogosphere, online articles have started to appear, e.g. George Ayittey's 'Africa through African eyes' (Abdou, 2010).

On the interaction between ICT and African culture, very limited documentation exists. There are just a very few scientific or philosophical works like Van Binsbergen's 'Can ICT belong in Africa, or is ICT owned by North Atlantic region?' (Binsbergen & Dijk, 2004), or Zakour's 'Cultural Differences and Information Technology Acceptance' (Zakour, 2004).

Orality

In his book 'Orality and Literature' Ong expands on differences in managing knowledge and verbalization in primary and oral cultures versus chirographic cultures. Ong notes that, for those acquainted with literate culture - by definition those that read an article like this - thought and its expression in oral culture appears strange and at times bizarre (Ong, 1982). Much of Ong's observations are diachronic, viewed through history, although in recent work he hints that further research on the differences between orality and literacy might produce new and interesting insights in interpretations (Bingham, nd).

There seem to be no (multi/trans)disciplinary models for science on the interaction with, and integration of, technologies in societies utilising orality as their main means of interaction. Although there is a persistent tendency among scholars that writing is the most basic form of language, Ferdinant de Saussure questions aspects of usefulness and shortcomings and dangers (Waterman, 1956). Changes in mental and social structures accredited to the use of writing have been documented (Goody, 1968). Ong mentions that even in the old days Plato already expressed serious reservations in the Phaedrus and his Seventh Letter about writing. He depicted it as a mechanical, inhuman way of processing knowledge, unresponsive to questions and destructive of memory (Ong, 1982).

Academic interventions encounter the complexity and vast schism between oral and literate culture. Obviously colored by conventional wisdom, orality is explicitly or implicitly attributed with a lack of introspectivity and analytical prowess (Ong, 1982). Although Sayed observes, 'Adaptation of technology and the becoming of a person are not separable conceptually from the evolution of the complex system that the community is' (Sayed, Singh, Saad-Sulonen, & Diaz, 2011), even academic interventions adhering to Complex Adaptive System methodology have not taken into account a measure of orality as a context-sensitive constraint.

Role of ICT

The role of ICT in advancing growth in Least Developed Countries is a hot topic of academic research and debate (Frediksson et al., 2010; IEG (Independent Evaluation Group), 2011; ITU, 2011a; ITU, 2011c; Toyama, 2010; Unwin, 2009; Zambrano & Seward, 2012). Regularly, attention is requested for the plight of the least connected, and institutes and nations are urged to collaborate in addressing the needs in rural areas, for instance during World Telecommunication and Information Society Day 2011 (WTISD) with it theme 'Better life in rural communities with ICTs' (Toure, 2011; UN, 2011).

Many countries collect Universal Service Funds to support ICT deployment in rural areas, specifically targeting growth of ICT services in rural areas (Calandro & Moyo, 2011). This body of work can be classified in the 'advocacy' category.

Over the past decade great progress in the availability of bandwidth improved Africa's connections to the rest of the world. Sea cables are floated to the continent. There is much growth in use of the mobile phone and social media like Mixt and Facebook. However, several obstacles have hindered ICT implementation and particularly impacted the adoption and utilization of ICT for the benefit rural communities. Integrating academic interventions in communities is a non-trivial, transdisciplinary, and utterly complex endeavor in rural Africa.

There is a growing understanding that a western-focused look at ICT is constrained. Researchers are beginning to question their precepts (Gomez & Pather, 2012). Anderson et al. found their measurements on 'a Digital Study Hall' in India was prone to conceptual and methodological difficulties (Anderson, Robertson, Nabi, Sahni & Setia, 2012). And a working paper on South African efforts to enhance the livelihoods of the rural poor through ICT concluded that only 'community-generated initiatives' have an opportunity to be sustainable, self-managed and an ability to be appropriated by local communities (InfoDev, 2008).

Digital Divide and Digital Exclusion

A wide variety of papers deal with the so-called Digital Divide. The various definitions of this divide converge on the notion that the digital divide refers to any inequalities between groups, broadly construed, in terms of access to, use of, or knowledge of, information and communication technologies (Servon, 2002). The messages in the literature convey multiple perspectives on the commonly agreed potential of a technology gap to exacerbate the already wide disparities between people in society (Unwin, 2013). Pais (2007a) proposes a mechanism for bridging the divide, based on the position that the unifying vision that telecommunications can empower people to meet their needs must be held by all stakeholders if communities are to value technology as a means of achieving sustainable prosperity. The digital divide is modulated by access, training and content; driven by market forces; and widened by unequal investment in infrastructure, discrimination, insufficient policy efforts, and culture and content (Servon, 2002). Further, the multi-dimentional nature of the digital divide features prominently in literature, citing the ability to access, adapt, and create knowledge using ICT to be critical to social inclusion (Warschauer, 2003). Warshauer explains that not only access is crucial, but that both technology and social developments play important roles while integrating technology into communities, institutions, and societies. He states that not so much the physical availability of computers nor the internet are most important but rather people's ability to make use of those technologies to engage in meaningful social practices.

Hudson made an attempt to investigate research, project and policy initiatives to examine the role of ICT in the transition from a rural village to a global village (Hudson, 2006). A significant number of United Nations studies, especially from its specialized agency for information and communication technologies, the International Telecommunications Union (ITU), show clear disparities in the availability of access and technology, still an important issue well into the 21st century (Frediksson et al., 2010; ITU, 2011c; ITU, 2010a; ITU, 2011a; ITU, 2010b; ITU-D, 2011). Also multi-national institutes like The World Bank publish cases in ICT, e.g. Baldwin & Thomas (2005). Likewise the Commonwealth Telecommunications Organisation (CTO) studies and reports on ICTs in the developing world (Calindi, Pulkkinen, Tongia, Akwetey & Ross, 2010).

Expanding the issue of internet access, a range of reports focus on the value and importance of broadband, including the issue of rural broadband in broad and generic terms only (Alves et al., 2011; Budde, 2011). Although the reports show an accelerating growth of capacity, ubiquity, and convergence (especially in the field of internet and mobile networks), this all seems to fuel a shift in the nature of the digital divide towards digital exclusion (Grosskurth, 2010). An array of reports, mostly from (People) Public Private Partnerships (PPPP), deal with derivatives of this issue, e.g. Carvalho, Klarsfeld & Lepicard (2011).

Connectivity and Labor

Little or no academic interventions address the cost of connectivity in rural areas in Africa, although this is an important denominator for the economic sustainability of internet networks in such environments. In general, little quantified data is available in the African telecommunications market. Available sources are ITU reports, the occasional demand-side releases, telecom operators annual reports, or market reports from commercial sources. Most notable is ITU's ICT Price Basket (IPB) which led the ITU to herald that ICT services are getting more affordable worldwide (ITU, 2011b). The report documents that nine out of the top 10 countries showing the greatest decrease in the ICT Price Basket value were from Africa. However, all of them had high values (i.e. high prices) to start with, and the report concludes that Africa continues to stand out in its relatively high prices, with fixed broadband internet access costing on average almost three times the monthly average per capita income. The ITU produced a supporting video to visualize this disparity (Statshot, 2012). With 15 African countries being landlocked (the highest number on any continent) the challenges of costs for crossing other countries with terrestrial wires are severe (Léautier, 2012). However, statistics lack focus, robust data does not exist, and methodological studies on communications costs in rural areas of Africa are virtually non-existent.

Access to telecommunication services such as the internet has a direct and mutual correlation with the gross domestic product (GDP) per capita of a country (Andrianaivo & Kpodar, 2011; Baldwin & Thomas, 2005; Gillwald, 2008). The literature suggests that telecommunications access strongly influences the financial and social well-being of a population. In contrast, the ITU reported in 2005 that the entire continent of Africa has fewer internet users than France alone (ITU, 2005). Current reports are more diffuse, due to convergence. They often aggregate the amount of mobile phone users. For instance, the Communications Authority of Kenya reports 4,716,977 internet users, of which 4,684,473 users connected via mobile phone during the second quarter of 2011 (CCK, 2011).

Although there are stark contrasts in terms of urbanization in Africa (United Nations, 2009), in developing countries, the vast majority of people (approximately 70% to 85% of the labor force) live in rural areas (Kozma, 2006) like Macha. In Zambia, 94% of its surface area is considered rural (Adams, 2003). 61% of Zambia's populace live in the rural areas (Central Statistics Office Zambia, 2011). For a developing country to increase its productivity and for the populace to enjoy an improved quality of life, it is essential that rural areas are developed to the extent that new opportunities are created and innovations occur (Chief Chikanta & Mweetwa, 2007). Chief Chikanta et al. mention that rural areas face significant resource challenges such as poor communications, transport, electricity and water supply. They deem a disparity in internet access is undesirable because it demonstrates that people of different communities have unequal opportunities to benefit from technology in their daily lives.

Further, there are no studies in rural areas on the cost of personnel in relation to their added value. Schwalje (2011) notes this discrepancy and provides some initial thoughts on concepts for national skills formation for knowledge-based economic development. In personal discussions, Schwalje confirmed the relevance for environments like those found in Africa, and thus the necessity to include data from Africa, but also the difficulty in finding and accessing relevant sources of data.

MANAGEMENT OBSERVATIONS

This section endeavors to explore management theories affecting views on how to perform and analyse operational internet access activities in rural Africa. House & Aditya (1997) substantiate findings that leadership literature mostly reflects Western industrialized culture. They conclude that 98% of literature is distinctly North American in character. Therefore this section wades through literature to unearth knowledge that could be relevant to review the management of internet initiatives like LinkNet in rural Africa.

Seminal Management Literature

Covey (2004), in 'Seven habits of highly effective people' identifies: 'first try to understand before you want to be understood.' Kotter in his book The Heart of Change advocates 'Commitment from the top and Show Quick wins' as key success factors (Kotter & Cohen, 2002). Senge et al. describe coherent theories in 'Presence' (Senge, Jaworski, Flowers, & Scharmer, 2011) and Charmer (2009) poses 'Theory U.' The idea is that it is important for fundamental change to have an open mind and truly listen to the essence. That means that one must invest time to find out what really matters. One listens not with the head but the heart, with the whole being. When the time of insight comes, action follows, smoothly and quickly.

Collins (2001), in his book, Good to Great , answers the question of why some organizations have achieved major transformation and others not. He poses the principle of 'first who , then what '. First one ensures relationship with the right people, then one decides where one goes. Most organizations act the other way around: first devise a strategy or a plan, then find the people. This does not work, postulates Collins. It must be: Who first, then What. Gratton (2007) in her book Hot Spots , deliberates on the reasons why some organizations or divisions are "hot." In such places there is a lot of action, happenings, inspiration, and achievement. One factor she mentioned is that a sense of urgency and sense of excitement is generated by asking questions. These are called igniting questions; questions or comments that made people stop and think, and then take action.

Other relevant thoughts can be found in Systems Theory. Examples include Ackoff's book, Redesigning the future: a systems approach to societal problems '(Ackoff, 1974). He indicates that a business cannot be separated from the system and the context in which it acts. Furthermore, discipline cannot solve complex business problems. Everything is interconnected and therefore important. The whole system is to gain an understanding. In his book Process Consulting , Schein (1988) deals with the question of what help really is. He shows the need to be open-minded and participate in helping those who need help. It is counter-productive to formulate many solutions. The crux is that those supported discover, self start, and deal directly with the challenges and opportunities, congruent with the concept of Social Innovation (Stam, 2012).

Integration of this seminal management literature with the positioning of academic intervention and the internet network activities in Macha is described through the Macha Works model (Matthee, Mweemba, Pais, Stam & Rijken, 2007; Stam & Oortmerssen, 2010).

Economic Growth

A growing number of reports expound that global business cannot afford to ignore the potential of the African markets (Roxburgh et al., 2010). Some exclaim 'It is time for Africa' (Otty & Sita, 2011). 'Africa's economies are amongst the fastest growing in the world, but growth does not necessarily lead to development' (Maczimbamuto-Ray, 2012). The heralding of the economic growth in an African environment is prone to complexity. Although Unwin postulates 'there will be no end to poverty' (Unwin, 2007), he recognizes the intrinsic richness of Africa (Unwin, 2008). Van Oortmerssen (2007) and Sachs (2011) both raise the issues of sustainable progress, taking into account the limits in resource use that the world has reached. Ndi (2010) indicates the consequence of the fixture on economic growth from an African perspective.

Development Innovation

There are calls to define the word 'development' for goal fixing. Provocative angles of viewing development come through a range of writers. Seminal are Freire's reports on his experience, and the way through, in situations of ignorance and lethargy which he deems to be a direct result of economic, social and political oppression (Freire, 2000). He poses a framework in education, pulling in actual experiences as learning cases, re-defining realities to allow people to be free to grow. Zambian citizen Moyo's book Dead Aid has been a landmark in the discussion of development aid (Moyo, 2009). The complexity of the issue can be seen through Millennium Development Goals committee chairman Sachs. He penned his thoughts upon reaching Lusaka, the capital of Zambia, for the first time '.. I knew that things were different, but I still did not fathom just how different' (Sachs, 2006).

ICT AND TECHNICAL ASPECTS

Literature on challenges of engineering in rural Africa is scattered over many academic fields. Further, reports on pilots or activities in remote or disenfranchised areas are also scattered over many fields. In this section we endeavor to gain insight into the status of ICT research relevant to the ICT access in an environment like Macha.

In the field of Information and Communication Technologies, the community of both technical and social science researchers in ICT4D (or ICTD) takes special interest in academic interventions on the African continent. Also in the area of computer science there are specific science teams operating, to look into technology-related activities related to rural Africa. However, ICT theories for Africa are being formed without any significant influence of African scholars due to unavailability of literature from African origin (Gitau et al., 2010). A recent study on the changing, interdisciplinary field of ICTD analysed 948 peer-reviewed academic papers that appeared in five peer-reviewed journals and two conference series between 2000 and 2010 (Gomez et al., 2012). In their conclusions, Gomez et al. deduce that the field is starting to help shape the design of novel technologies for developing world contexts.

Grosskurth (2010) exposes a positive outlook on technology in Africa, while UNESCO's landmark report, 'Engineering' provides insight into the particulars of the shortages in Engineering (UNESCO, 2010). UNESCO mentions that 2.5 million new engineers are needed in the African context (IEEE, 2012). Doing research in developing regions like certain countries in Africa are blighted by challenges (Brewer et al., 2006), and to keep rural wireless networks alive beyond a pilot stage is another recognized challenge for academic interventions (Surana et al., 2008).

Wireless Networks

Only a small number of rural wireless internet networks are documented in Africa. In his thesis, Nungu (2011) describes one in Tanzania. Another example is the Peebles Valley Wireless network in South Africa (Johnson, 2007) which could be most closely compared with the internet project in Macha (Matthee et al., 2007). These networks are unique in that they overcome, and document, challenges like long distances between wireless nodes, low-bandwidth gateways to the internet, lack of reliable power, and high cost of internet connectivity. They usually share a low-bandwidth, costly link to the internet amongst a large user-base. The adagio is that analysing and understanding the traffic distribution, web usage patterns and source of bottlenecks can facilitate network designs that are optimized to give rural users a better internet experience and bring down usage costs. Further, they operate in a mixed, and complex environment, with multiple goals as in development and research.

Internet Traffic

A large multi-year study of the distribution of internet traffic in the developed world was carried out by Ipoque in 2008/2009 (Schulze & Mochalski, 2009). It showed that P2P traffic had decreased significantly and Web traffic had increased due to the extensive use of content servers. In Germany, for example, P2P traffic decreased from 69% to 52% and Web traffic rose from 14% to 26%. Johnson, Pejovic, Belding & Stam (2012) consider that the work that could be closest to the rural Zambian context is one that examines Web traffic usage in internet Cafes and community centres in Cambodia and Ghana (Du, Demmer, & Brewer, 2006). Here only HTTP traffic was studied, without a wireless network aggregating traffic to the internet connection. Classification of traffic showed that sites like Yahoo and MSN, advertisement sites and multi-media content accounted for the bulk of the traffic.

The performance of rural area wireless networks has been investigated since their inception in the early 2000s. Surana et al. (2008) present a comprehensive study of rural area network problems. Here technical issues were juxtaposed with the social obstacles of deploying networks in the developing world, with a focus on system troubleshooting and an accent on energy related problems. In addition, Surana et al. provide anecdotal evidence of the social problems. Specifics of network usage in the developing world is also the subject of Sen, Kole & Raman (2006) and Kumar & Best (2007). The former investigates a specific application (VoIP) and its economic feasibility in rural areas, while the latter includes interviews of telecenter users in rural India. Wireless network performance in rural networks have been analysed (Bhagwat, Raman, & Sanghi, 2004; Surana et al., 2008). The academic intervention through the Peebles Valley mesh network in South Africa showed that in an unloaded network, a mesh network becomes the bottleneck, rather than the satellite, after 7 hops (Johnson, 2007).

Johnson et al. used network traces from the internet network at Macha to explore the degree of local user-to-user interaction in the village. Social graphing, using instant message interactions on Facebook, revealed that 54% of the messages are between local users in the village (Johnson, Belding, & Stam, 2012).

Viruses

Experience in the internet network in Macha signals the important issue of virus traffic, especially in the highly constrained bandwidth environments. In 2007, around 600 million machines connected to the internet. Of these, Goebel et al. (Goebel, Holz, & Willems, 2007) quote Vint Cerf's estimate of a quarter of them being infected with botnets. Others estimated the range to be between 12 and 70 million at that time. Goebel's study at Aachen University in Germany in 2007 showed that in 8 weeks, 13.4 million successful exploits were discovered due to 2,034 unique malware binaries circulating amongst 16,000 unique IP addresses. Botnets, which are the key platform for most internet attacks, are surveyed by Gu, Perdisci, Zhang & Lee (2008). This work highlights that botnet command and control has moved beyond the common methods such as contacting an internet Relay Chat (IRC) server. Current methods make use of HTTP and P2P traffic and are thus harder to detect with common intrusion detection systems. This emphasizes the fact that detected malware may only be the tip of the iceberg. The severity of the problem is clear from Bhattacharya & Thies (2011). Here experiences, behaviors, and unmet needs of telecenter owners are studied, with respect to their attempts to prevent virus infections on their machines. Virus control is recognized as a largely unsolved problem. The importance of user education in rural networks is highlighted by Ishmael, Bury, Pezaros & Race (2008). Uniquely, Johnson et al. analyzed full TCP dumps and provided insights on virus and bot presence in the internet network in Macha (Johnson, Pejovic, Belding, & Stam, 2011), for the first time in rural Africa.

Use of ICT

The internet has evolved both in terms of size and application since its birth. Recent studies have shown that the average web page size in 2012 was 68 times larger than the average size in 1995. It was 14.12K in 1995 (Domenech, Pont, Sahuquillo, & Gil, 2007) and 968K in 2012 (Souders, 2012). There is an increasing amount of off-personal computer (PC) storage and processing using cloud computing for services such as navigation, photo sharing and file hosting. Many applications that in the past were run on a user's device such as email, word processors and instant message clients, are now run on web browsers. These features have brought users in developed countries closer to the vision of 'anywhere any-time' computing, where devices connected to high speed Internet connections delegate computing power and storage capacity to cloud computing services. Postill (2008) highlights the fact that as the internet continues to grow, it is becoming 'more local' (Postill, 2008). This phenomena is beginning to blur the boundaries between online and offline social domains, and it is this trend that justifies a localization approach to network design, especially in isolated rural communities. However, no literature could be found on the implications for users in rural Zambia where internet access speeds of 64 kbps, up to 256 kbps, and latency over 700 ms are most common.

There are many studies on social network interactions on Facebook, both at a structural level using friend lists and at an interaction level using wall posts, e.g. Kumar, Novak & Tomkins (2006), and Mislove, Marcon, Gummadi, Druschel & Bhattacharjee (2007). Wilson et al. argue that social links created by friend lists are not valid indicators of user interactions (Wilson, Boe, Sala, Puttaswamy, & Zhao, 2009). This is shown by the fact that the number of 'friend adds' account for 45% of the activity per day whereas comments only account for 10% of the activity.

The lack of local content in rural Africa has been documented by Van Hoorik and Mweetwa (Hoorik & Mweetwa, 2008). According to their observations, rural Africans do not find a representation of their customs and culture online, thus they may perceive the internet as a foreign body . Manschot explores models of online content generation (Manschot & Stroek, 2009). The authors conclude that social and cultural factors have to be considered for successful implementation of 'content generation tools'. Locality of interest has been primarily studied in the domain of peer-to-peer networks. For example, a semantic clustering technique is employed by Handurukande, Kermarrec, Le Fessant, & Massoulié (2004). The semantic relationship is either implicit, using information such as peer-history, or explicit, using meta-information about the file, such as whether it is music or video.

Time-delayed proxy for bandwidth-limited networks was proposed in 2006 by Du et al. (2006). Vithinage et al. implemented and documented proxy behaviour (Vithanage & Atukorale, 2011). Locality of online interactions, on the other hand, was observed with the advent of online social networks. Wittie et al. exploit the locality of interest in order to improve online social networks' usability for remote areas (Wittie, Pejovic, Deek, Almeroth, & Zhao, 2010). Their work is geared towards whole countries, and thus large geographical areas. In view of this, Johnson et al. proposed an intervention by means of a upstream proxy in the internet network at Macha (Johnson, Pejovic, et al., 2012).

APPLICATIONS

In this section we search for knowledge on use of ICT in rural areas. Especially in this area, literature becomes thin. However, this is an important area, as Oudshoorn signals in 'How Users Matter' (Oudshoorn & Pinch, 2005).

E-learning

In recent years, much research has been devoted to the topic of distance education. Bernard deals with the comparison with classroom education (Bernard et al., 2004), and Garrison with its potential transformative effects (Garrison & Anderson, 2003). The concept of e-learning in resource-limited rural environments has also been explored; Pais (2007b), Hoorik & Mweetwa (2008), and Vallis et al. (2012) documented their application and implementation within rural Zambia. Chen, in his book Education Nation , describes the results in western schools that integrate modern digital tools to facilitate thinking, communicating, collaborating (Chen, 2010).

The field of distance learning pertaining to music is very new. The substantial cost of internet connection and video conferencing hardware was prohibitive to widespread synchronous education in music (Litterst, 2007). More current research has demonstrated that distance learning in music is not only feasible, but also functional on a basic level (Dammers, 2009). In Macha, a longitudinal ePiano project, music education through the internet, has been progressing for a number of years (Shoemaker & Stam, 2010).

Schurgers et al. expand on experience and reflect on the opportunities and challenges of E-learning in health in Zambia (Schurgers, Stam, Banda, & Labib, 2009).

Cultural Heritage

Preservation of intangible cultural heritage is part of healthy and sustainable progress of rural communities (UNESCO, 2003). Affiliated with the LinkNet activities in Macha, the local radio provides a platform for dissemination of cultural knowledge and wisdom in a community that operates mainly in primary orality. Programs provide a beacon for identity in a community in flux in a fast changing world (Stam & Mweetwa, 2012).

DISCUSSION

Nine years of ethnographic research in rural Zambia provided a lot of insights into challenges or the relevance of academic interventions and their applicability in the rural context. There appear to be a lot of discrepancies between western knowledge and its application in the disenfranchised context. Frameworks can be culturally inappropriate and locally impractical. Discrepancies include:

1. non-alignment of locally relevant topics with available (international) research
2. lack of accessibility of research findings in varied culturally-adapt manners
3. issues of language (English versus local language)
4. representation of knowledge, e.g. in literal format versus an oral format
5. relationality, e.g. writings by foreign research versus dissemination by local research
6. cultural positioning, e.g. linear categorization versus mediatory positioning
7. contextuality, e.g. described for an audience referencing to the resource-abundant environment versus a local audience referencing to a resource-limited environment
8. validation, e.g. international-peer-reviewed versus local-community-reviewed.

The almost exclusive discursive way of presenting results in ICT is providing a huge barrier to its use in other regions of the world, where discursive thought is scarce. Current academic discourse has little connection with realities and topics of interest in developing countries. Such topics center on, for instance:

1. people
2. analysis of effects of interventions in social relationships
3. the empowerment of individuals or communities
4. sense of accomplishment and feelings of being part of the world
5. effects of interventions to the social cohesion of the community
6. opportunities to strengthen local culture and archive events in the past
7. abilities to interact as a community at a collective level.

Key issues are agency, security and peace. Results of this misalignment fuels inaccurate and misleading stories in disenfranchised areas. This, in turn, discourages the desire to collaborate with academics. Only through the local expression of the intervention can the real embrace of the findings by the local community possibly take place.

There is no established scientific community in ICT or engineering in most disenfranchised regions of the world. Further, engineering research mainly takes place in the western environment, and financing focuses on such Western based research. No significant funding mechanisms for science flourish in Zambia. In collaborations, African, rural partners are often excluded from funding.

CONCLUSIONS

Much research in engineering incorporates a quantitative and technological perspective. Literature is biased towards quantifiable responses, aims at prioritization of needs and is focused on Western topics and conceptualization of solutions. Works often lack long term contextual evidence. In positivistic, technical sciences there seems to be little regards for culture and context. There is no localized science-reporting in most developing countries, and certainly little reporting in the local languages. All this hampers communities from being able to learn about academic interventions in a meaningful way.

There appears to be little regard for the potential difference in meaning of the technical artefact in a different contexts. Academic reports invariably focus on tangible, quantifiable and instrumental impacts following Western methodology. The lack of long term, longitudinal research on the use of technologies in disenfranchised areas particularly affects the knowledge base. Not many projects extend over multiple years or beyond the project implementation phase. Further, the intangible and unquantifiable results of interventions receive little attention. Literature is aimed at a generalist, mostly macro-economic altitude.

The discursive expressions of scientific knowledge, reduced to abstractions in English texts, seem to have little discernible effect on - or even defacto prohibits the inclusion of - oral societies. The foreign academic appropriation of local information for private, foreign profit, alienates the local community and renders it objectified and possibly exploited.

There are hardly any engineering studies on activities and results in disenfranchised areas from the perspective of disenfranchised areas. Reports base conclusions on short periods of observation. Validating the research with those in the areas under review hardly ever results in participation by, and/or recognition of the efforts and their worth, by local people.

There is a clear need for conceptual frameworks applicable to context and culture to be in equilibrium with local culture and heritage, taking into account locally important aspirations and paradigms. There is much room for locally developed models and locally enshrined research, performed on the basis of local tradition, possibly involving local indigenous institutions to assure good grounding, especially when dealing with rural communities.

ACKNOWLEDGEMENTS

We thank the community and stakeholders in Macha in participating in the study, and the staff of Macha Works who assisted with the study. We thank the reviewers for their valuable comments. We thank researchers David Johnson, Veljko Pejovic, Elizabeth Belding and Lisa Parks for the University of California, Santa Barbara and Tony Robert for Royal Holloway, University of London for their contributions and suggestions.

REFERENCES