No abstract

The computing for development community knows that technology interventions involve consideration of social, technical and environmental factors. Research into WiFi solutions has fallen off as ubiquitous mobile solutions penetrate even the deepest rural communities worldwide. This paper argues that the latest wave of WiFi mesh networks offers benefits that traditional top-down WiFi and mobile networks do not. In addition, we propose ethnographic and participatory methods to aid the effective rollout of mesh inverse infrastructure with and for a given community. This paper describes and then analyzes a mesh for voice rollout within a situated context. We explain how to conduct informed community co-design and how to factor in local socio-political concerns that can impact on the design, rollout and subsequent maintenance of community-based wireless mesh networks. While we have not yet analyzed baseline and initial usage data, we do have new lessons to offer.

Abstract. Designing for oral users in economically poor places has intensified efforts to develop platforms for asynchronous voice. Often these aim to assist users in rural areas where literacy is lowest, but there are few empirical studies and design tends to be oriented by theory that contrasts the mental functions of oral and literate users, rather than by local practices in social situations. We describe designing an Audio Repository (AR) based on practices, priorities and phone-use in rural Africa. The AR enables users to record, store and share voice files on a shared tablet and via their own cell-phones. We deployed the AR for 10 months in rural Africa and illiterate elders, who have few ways to use free or low-cost phone services, used it to record meetings. Use of, and interactions with, the AR informed the design of a new prototype. They also sensitized us to qualities of collective sense-making that can inspire new interactions but that guidelines for oral users overlook; such as the fusion of meaning and sound and the tuning of speech and bodily movement. Thus, we claim that situating design in local ways of saying enriches the potential for asynchronous voice.Keywords: Oral users, Rural Africa, Asynchronous voice, Social media. IntroductionMost of the world's 800 million illiterate adults live in economically poor regions; in fact, interpolating Worldbank indicators, in Africa fewer people can read and write than there are cell-phone subscriptions [23]. Cell-phones offer many new opportunities to people in technology-sparse regions but they may also amplify gaps between oral and literate users and this has motivated recent efforts to develop asynchronous voice platforms for rural settings where literacy is lowest. Such efforts include interactive voice applications and services for cell-phones to record and access content produced locally, or by authorities or NGOs [19,27,31,39] or participate in citizen journalism or online social networks [25,29]. However, there are few empirical studies [29] and design tends to be oriented by theory that contrasts the mental functions of oral and literate people, not by practices in social situations. In this paper we describe a system we deployed in rural South Africa to enable users to record, store and share voice files for free. We aim to show that situating design in local ways of saying can inspire new interactions for asynchronous voice; thus, we start by claiming that theory currently orienting design for oral users omits factors that can prompt ideation.

We reflect on long trials of two prototype social media systems in rural South Africa and their biases towards certain communication practices on information sharing. We designed the systems to assist people in low-income communities to share locally relevant information. Both involve communal displays, to record, store and share media, and users can transfer media between the display and their cell-phones. MXShare, which we report for the first time, also enables real-time, text-based chat but AR enables sharing only audio files asynchronously. Both systems were located at the same sites for community communication and co-present oral practices effected media recording and sharing. Their use reinforced differentiations in sharing information between older and younger people. We argue that designing social media systems to widen information access must respond to complex interactions between social structures and genres of communication.

We consider practices that sustain social and physical environments beyond those dominating sustainable HCI discourse. We describe links between walking, sociality, and using resources in a case study of community-based, solar, cellphone charging in villages in South Africa's Eastern Cape. Like 360 million rural Africans, inhabitants of these villages are poor and, like 25% and 92% of the world, respectively, do not have domestic electricity or own motor vehicles. We describe nine practices in using the charging stations we deployed. We recorded 700 people using the stations, over a year, some regularly. We suggest that the way we frame practices limits insights about them, and consider various routines in using and sharing local resources to discover relations that might also feature in charging. Specifically, walking interconnects routines in using, storing, sharing and sustaining resources, and contributes to knowing, feeling, wanting and avoiding as well as to different aspects of sociality, social order and perspectives on sustainability. Along the way, bodies acquire literacies that make certain relationalities legible. Our study shows we cannot assert what sustainable practice means a priori and, further, that detaching practices from bodies and their paths limits solutions, at least in rural Africa. Thus, we advocate a more “alongly” integrated approach to data about practices.