Uganda hosts the largest population of refugees in Africa, and has a progressive ‘open-door’ refugee policy. Uganda’s refugee settlements are not fenced-in camps. They constitute huge tracts of land,¹ in which refugees are encouraged to farm, make a living and trade with hosting communities (see Figure 9.1). Many of these settlements contain ‘old caseload’² from as far back as the 1950s, from events such as the ‘Acholi Wars’ and the Rwandan genocide. In settlements near remote borders, arriving ‘new caseload’ refugees are typically instructed to go to certain hillsides in huge designated areas and settle on a plot 50 metres square. These areas are – by definition – just wilderness, and for decades refugees have been settling in this way relatively peacefully. So nobody thought to record where they were.

Documentation: the problem of population

In 2016, violence in South Sudan created a sudden mass influx of refugees across the border into Northern Uganda. The world looked on as the refugee coordination responded, adapting areas and opening new settlements. One of these – BidiBidi – became famous at that time as the largest refugee settlement in the world. With this event came the realisation that decision-making in both the United Nations High Commissioner for Refugees (UNHCR) and Uganda’s Office of the Prime Minister (OPM) were under-informed. Of the 500,000 or so new arrivals, who knew what communities were where in relation to water, medicine, sanitation, education and security? These issues were exacerbated as another mass refugee influx followed less than six months later from the west (the Democratic Republic of Congo).

By January 2018 refugee figures were reported at 1.4 million in Uganda, but this was still an estimate. It was becoming increasingly acknowledged that, despite decision-makers having said that there was some state-of-the-art interventional technology at their disposal, they were not sure how it worked and thus how it could be adapted to the increasing complexity of the situation. In the capital (and even field offices) nobody seemed to know what was really out there. In February 2018 several officials were suspended over what arguably amounted to a long-term lack of engagement with data (Okiror, 2018), as some of the financial discrepancies between funding and delivery started to become undeniable.

In response, by late 2018, a registration of refugee individuals, using biometrics and other top-down digital methods, was published by the UNHCR/OPM response coordination (UNHCR/OPM, 2018). Authorities could now approach a realistic estimate of refugee numbers in Uganda (at the time of writing, officially 1.2 million). But evidence suggested that this technique of enumeration was only partially useful. The ambiguities of the term ‘hosting’ are geographically complicated: Are internally displaced persons (IDPs) refugees? What if refugees had been born in Ugandan settlements? The multiple variables of cultural/gender, diverse medical needs and compounded geographical unknowns within and without settlements still made it almost impossible to allocate resources, stop disease outbreaks and control intertribal or international violence in the districts hosting incomers. And so the old problem of the ‘last mile’ (Balcik et al., 2008) remained obscured, as designated resources still failed to actually arrive on the terms under which they were originally provisioned.

UN sector challenges: formal and informal infrastructure

United Nations (UN) bodies carefully merge their systems as much as possible with national government, but humanitarian coordination over the decades has also become increasingly sector led. These sectors are traditionally split: water, sanitation and hygiene are dubbed ‘WaSH’, hospitals are ‘health’ and schools are ‘education’. Sectors such as ‘energy and environment’ may at first seem non-emergency but disease and famine are immediate when unregulated deforestation, for example, causes populations to run out of wood that is much needed for sanitation and cooking. Equally, market and supermarket locations (cash-based intervention – CBI) have been notoriously vital for emergency response in sudden-onset disasters such as earthquakes or tsunamis. That police outposts fall under ‘social amenities’ but also under ‘protection’ (a sector that touches on all other sectors) demonstrates how this framework is in itself a work in progress.

The concept of a complete set of statistics on refugees implies many assumptions (not least the fundamental anomaly that statistics describe a pursuit of ‘stasis’, and migrants are inherently mobile). For example, refugees are sometimes rich, they frequently return home and often they do not live in settlements where they are registered, let alone on their plot. Refugees may try to move house, often do business outside of settlements or even outside Uganda.³ In other words, they are human. To this day in West Nile, the situated challenge of the human factor prevails and as the emergency response becomes one of development, authorities in advanced response modes still lack sufficient intervention information.

The accumulating mass of humanitarian institutional growth has made it difficult to operate holistically across or outside these sector frameworks and accommodate refugee complexities. The UNHCR established the Comprehensive Refugee Response Coordination (CRRF) to assist in the coordination of better inter-sector communication.

The digital landscape in Uganda

Historically, there has been a lot of data gathered in Uganda to try to tackle the issue of refugee management: information on people’s needs, access to food, water and medicine. Because of the geographical and financial scale of the response, this has, almost exclusively, been sampled qualitative data from key informants representing scattered community views. Due to sheer numbers, Uganda has long been a global case study for the challenges and successes of refugee management. When the Northern Uganda influx peaked, the holes in this data became a part of the crisis.

After the mass influx, it was clear that much of the data being generated were not capable of accommodating important home-grown local priorities. Informal micro-resources, such as thousands of small rural wells and village doctors, were not being compared with official (and expensive) geospatial records gathered by scores of different ‘implementing partners’, non-governmental organisations (NGOs) which worked under the refugee coordination. Even if UNHCR were able to curate all the partners’ data on their interventions, this data, gathered under multiple standards and partitioned by sector, could never prevent bore holes from being dug in areas already full of natural wells or health centres from being located too far from populations in need. In short, resource allocation was a fragmented mess unless it could incorporate empirical, comprehensive, up-to-date information on everyday community resources.

Data siloing and proprietary data: Plug and Play

The humanitarian sector creates a sometimes fetishistic drive for innovation. Uganda is inundated with well-meaning technological solutions, and these include digital information collection and data management projects. Both NGOs and UN agencies have the unfortunate habit of hiring independent commercial (for-profit) consultants for tech-enabled interventions, rather than allotting the time of their own staff. Reasons are complex, but include heavy workload and lack of in-house capacity-building methods.

Another problem is that funding and support is predicated on (often digital) innovation, so hard-pressed humanitarian sector organisations might choose to pursue technology over social engagement. For many NGOs, innovation around data can result in information becoming locked or siloed within proprietary software because it is often in the interests of the consultants to restrict or deny access to content (and even methodology). With an eye on cutting costs, very few NGO-contractor terms of reference stipulate extended or indefinite sharing of data (e.g. uploading information to technically accessible open platforms) at the end of these projects. The resulting data sets are sometimes declared ‘open’, but may stipulate that the data must be accessed through a controlled private website. The passive mode of openness here is effectively artificial, because it compromises access through systemic exclusion (conditions on sharing/registration of other data, for example).

Governing authorities in Uganda believe that the way to sustain effective humanitarian intervention is to initiate a handover to local and national authorities. In some instances, this is made more difficult by the constant push for aid-linked innovation. Rather, funding for internal system reinforcement might prove more helpful where the skill of local authorities also faces practical obstacles enshrined in a history of non-adaptive improvements, which still proselytize ageing or obsolete technology environments: network electricity-dependent workstations; software licences with ‘use-by’ dates, and ‘helplines’ rather than global community support; software licences that need maintaining; electricity needs to be reliable for desktop computers; traditional office systems have been set up, under which staff are dependent on commercial helplines for support.

Innovation overload

Across the world, the responsibility upon everybody to engage with data is becoming increasingly inevitable, even if simply for self-protection. Whether the world likes it or not, it seems that avoiding the demands of the digital can have worrying implications. We love ‘plug and play’ products, and as Tim Ingold writes (2011, p. 26), it is

New humanitarian products are frequently remotely conceived, initiated and deployed by a ‘telepresent’ and commercially interested society. Digitally enabled entrepreneurs serving the humanitarian sectors usually work from remote and insulated positions. Digital humanitarian solutions are sometimes created with little conception of (or interest in) the complexities of field-centred ontology. Further, through constraints of aptitude or time, there often seems to be a failure by responders to engage with the workings of these technological products and the communities on the ground. The speculative nature of some designs can even create a situation of need where there was none before. In precarious settings, inappropriate but well-meaning innovations can be at best ‘litter’ and at worst even life-threatening. The unused software and data in this landscape could start to resemble the obscene and redundant surplus hardware from notorious ill-judged humanitarian interventions (like the eventually over-provisioned ‘Live-Aid’ project of 1985, whose trucks and planes still litter the runway at Lokichoggio supply airstrip in northern Kenya). But this ‘useless litter’ is sometimes data, sensitive data. Worldwide, technological hardware continues to emerge, but with decreasing ‘hard-wired’ improvement between generations. Nowadays, accessible technology in sub-Saharan Africa is arguably matching this ceiling and it is often amply fit for purpose.

The expectation for ‘plug and play’ solutions can create a disinclination to understand the actual components of a problem, or attend to the semantics of site-specific issues – the ‘ink’ of the intervention. Granular-level data are demanding and difficult to deal with, but if the labour of engagement is not applied to technology – by commissioning decision-makers as well as technicians – the life-saving imperative to make the humanitarian landscape more accountable could degenerate into even deeper states of disconnectivity than ever before.

Technological ethics

There is growing ethical discussion of ‘the digital’ in humanitarian action, and mapping in Africa has always been a techno-colonial exercise (Dirk et al., 1996). Enthusiastic mappers declaring a commitment to map every corner of Africa should beware of the adage that ‘maps are never value-free images’ (Cosgrove and Daniels, 1989, p. 278). Critical writers from David Harvey (2009) to Edward Said (1978) clearly make the case that when we map there are preconditions embodied within the process. The exploitation of personal data is well reported. On African consumption of technology, Nanjala Nyabola (2016, p. 158) writes:

Lessons can be learned, too, from other reports of self-interested participations in ‘publicised’ data transparency, as can be seen from D’Angelo and Ranalli’s (2019) critical view of information politics in and around the US electoral process. Freedom of information and transparent data access can potentially evolve into unethical commercialisation of information, which can damage the best interests of democracy.⁴

Nevertheless, geographic information systems (GIS) are a vital tool when dealing with dispersed populations in need: digital technologies generally enable NGOs to outstrip previous reporting frameworks in terms of accuracy, granularity and field intimacy. With the mention of ‘technology’ and ‘data’, many humanitarian donors are satisfied with the promise of clearly improved intervention feedback. But many of these data still only partially represent the wealth of community engagement and field accountability that could be available.

HOT pilot: crowdsourcing non-settlement refugee data in the West Nile

In certain ways, the Humanitarian OpenStreetMap Team (HOT) could claim to be different. Important open-source ethos holds that important apps and software for managing information in resource-poor settings have been available (for free) for many years. And the ethos accompanying the technology is fully, and inclusively ‘non-profit’. And free. For OpenStreetMap (OSM), this is in the form of Maps. Me, OSMAnd, QGIS and ODK (Open Data Kit)/Kobo. So this makes the problem one of better application of basic digital tools, rather than the introduction of new technologies: adaptation, rather than innovation. Previous socio-economic blockers to digital inclusion are quickly dissolving with the market proliferation of significantly powerful and available smartphone handsets. Therefore in Uganda a HOT intervention seemed highly feasible, but it needed to be different from other data projects.

A funding call was posted, and the relevance of the open-data ethos to the situation (and to the UN Sustainable Development Goals (SDGs)⁵ themselves) was recognised by the US State Department funders. So in March 2017, HOT initiated a pilot project in Northern Uganda titled ‘Crowdsourcing Non-Camp Refugee Data’, with a sister project in Istanbul. In Uganda, this became specifically framed to address the challenges that the UNHCR identified: resource allocation, host-community engagement, community inclusion and affordable – yet sustainable – information management. The project was to experiment with capacity-building using community-owned smartphones, crowdsourced data and open-source GIS. Although very different, both the Uganda and Istanbul projects proposed to expose self-identified community needs in precarious populations, and in Uganda, to present a ‘common operational picture’ to UNHCR (Wolbers and Boersma, 2013). HOT, whose short but avid history had encountered equivalent crises, reckoned that inclusion of previously undocumented ‘informal community resources’ would provide improved understanding of vital infrastructural realities on the ground.

Open-data format works across, within and without sectors, but working from the bottom up is very different in nature from the compartmental ‘sectored’ traditions of humanitarian information management. Mapping teams uniquely do not conduct the surveys themselves, rather they teach OSM methodologies to community members, who themselves become the surveyors. Donors and responders are then increasingly able to see life-saving data changes clearly pictured on OSM. The potential for the allocation of essential resources per refugee family, even per capita, is compelling. With everybody collaborating and working together, mapping both inside and outside of settlements, this could also be a way to address issues of inequality and conflict between refugee and host communities. As an independent self-organising movement,⁶ HOT could see a solution in Uganda that existed independently from the restrictive legislation of top-down ‘owned’ information. Once funded, HOT was able to citizen-activate mapping teams, disregard the constraints of information politics and, based on nothing more than a fundamental right to internet access, start collecting data. Community-mobilised ‘motorcycle mapping’ techniques (which had proven effective in the 2014–16 Ebola outbreak in West Africa)⁷ would play a central role in this intervention.

OSM origins and history

HOT’s global agenda is characterised by the (self-)representation of underserved and hidden communities. It focuses on empowering people within their own communities to take control of how they are represented, mapped and seen by the outside world, so the authorship and ownership of data remains with – and can be used by – the participants themselves. Communities can be empowered to use this citizen-generated GIS as a way to have the local voice heard. At the same time as joining a socially cohesive global and local movement of OSM, they can also use the platform for small business and personal navigation (like a more complete, publicly authored version of Google Maps). Importantly, this model is sustainable and free. Intrinsic to it is the creation of a local workforce that could, in this instance, join in on implementing and maintaining data themselves. And so, in these sudden influx conditions in Northern Uganda, this data project, initiated as a pilot, only a ‘proof of concept’ at the time, rapidly became the source of high-quality, actionable information for emergency use.

OSM can be imagined as a global creative collaboration platform, a publicly authored ‘wiki map’. The project finds its heart not in the technology or tools it uses, or commercially interested organisations backing it, but in the OSM map itself. It can also work offline. OSM launched in 2004, when a 24-year-old entrepreneur, Steve Coast, set up ‘a collaborative project to create a free editable map of the world’ (Wikipedia, 2019b), inspired by the success of Wikipedia and the predominance of proprietary map data in the United Kingdom and elsewhere. Since then, it has grown to over two million registered users, who can collect data using manual survey, global positioning system (GPS) devices, aerial photography and other free sources.

OSM can record and geolocate images, sounds, web links, videos and all map data, and is free to use under community principles of mutual inclusion. It is supported in the online learning environment at https://learnosm.org and the global community platform https://www.wiki.osm.org. This resource (which in March 2018 recorded its 30 millionth edit) has for some years been harnessed by the global humanitarian community to assist in connecting community needs with resources in both response and development settings around the world. It was first notably used by humanitarians in Haiti in 2006, and again in the 2010 earthquake, in order to rapidly create accurate and current digital maps showing critical factors affecting the local population. During this and subsequent years, a humanitarian OSM organisation dedicated to creating ‘fast-cycle’ missing maps (Johnson et al., 2010) for humanitarian emergencies was conceived. The HOT/Missing Maps community worldwide is now made up of over 100,000 humanitarian OSM practitioners (see Figure 9.2).

As one of the three founders of the Missing Maps project with the American Red Cross (ARC) and Médecins sans frontières (MSF), HOT has been making and delivering geospatial digital (and paper) maps to assist in response to humanitarian crises since 2010. This humanitarian OSM movement can be seen as both a resource, a methodology and a philosophy. HOT shifts its shape to suit different contexts and the humanitarian OSM team has bilateral modes of distributed community network. A blank map continues to evolve, as remote mappers gather at group ‘mapathon’ events to focus on disaster sites by using field-informed crisis mapping tasks. These are posted and coordinated on the HOT online ‘tasking manager’ and contributors select individual grid squares to hand-digitise and outline buildings, roads and infrastructure (e.g. hospitals, schools, markets). The OSM-user communities on the ground download the map, add values to the points (features), lines (roads/rivers) and shapes (fields, swamps, forest). They use open-source smartphone apps and other analogue tools, such as hard-copy annotated maps,⁸ to record and reupload this information.

Reception of these methods by government authorities and NGOs

Painstaking efforts have been made in Uganda by UN coordination to account for partners’ intentions (i.e. commitments to help by humanitarian coordinators), but in 2017, the picture of field realities and activities on the ground was fragmented. Aspects of refugee life were clearly being missed by the traditional UNHCR approach, as witnessed by HOT field surveyors in the field, who, once deployed, were repeatedly met with surprise: NGO staff had often never been seen in plots lying more than a short distance from major supply roads – not only in the surrounding areas, but inside the refugee settlements.

Uganda community mapping was to yield impressive results. In the first 12 weeks of surveying West Nile, ten HOT-supervised community surveyors recorded 8,300 public amenities in the Arua District, as well as community verifying the few existing geotagged features. This took only two field staff and two managers to coordinate and the performance was further replicated and scaled into other districts. The cost, in terms of humanitarian budget, was minimal. The investment value in useful data points was huge. From this intervention, information on many features could be analysed and local authorities were to report major engagements with the data for resource allocation at district, sub-county and village level.

Even simple maps at this level of detail proved arresting (see Figure 9.3). By colouring nodes (surveyed points on the map) at which the answer to ‘Is this functional?’ was ‘No’, around 30 per cent of the 3,500 water features mapped in the first two months could visually be identified as non-functional.

Once the community had been mobilised en masse, this type of visualisation was simple to produce. But when these maps were first shown to UN sector leads they were met with mistrust. A common reaction to the ‘open’ part of OSM is a fear that all data will be made public: ‘Who had collected the data?’ ‘How could they be trusted?’ ‘How come they show so much more data than “official” maps?’ and most of all, ‘These “non-functional” reports can’t be correct!’ But HOT’s information not only confirmed all existing UNHCR data, but additions, updates and corrections were enthusiastically corroborated by UNHCR field officers on the ground.

Community data, open data

Michal Givoni (2016, p. 1025) describes crisis mapping as ‘a new modality of participatory humanitarian action in which global publics are mobilised to trace digital maps of disaster-stricken sites and to classify, verify, and plot on maps big data produced by disaster-affected people’. Participatory mapping enables cheap and accurate resource allocation and data de-siloing, mainly because of the time invested by the international community. It has been possible for some years to ensure that data on individuals (e.g. consented medical details) never leave the hard drives of the partner implementing the surveying. But demographic data (such as population numbers per site) can be shared and downloaded via the Humanitarian Data Exchange (HDX) by humanitarian organisations registered and vetted by the United Nations Office for the Coordination of Humanitarian Affairs (UNOCHA). From this, OSM uploads can be chosen with precision at any stage in the process. But these ‘commits’ should always acknowledge accompanying responsibilities towards ongoing data stewardship of currency, ownership and social relevance.

The concept of openness of data is better explained by the concept of open and complete control. No third parties mediate the data collected, and the user is never compromised in this. But with total control comes total responsibility. The will for this existed in Uganda, but it needed to be matched by specific commitment to engage with fast-evolving changes in global concepts of formal validation (and the growing acceptance of movements like OSM). OSM is open and editable by those engaging with it and is governed by the self-policing community. OSM does not have the potential to be exploited for mass marketing because, under the Open Database Licence (ODbL) laws, those exploitative bodies are required to reciprocally share all of their own data if they use open data. OSM is systemically protected by this caveat. However, that does not necessarily make digital recording of populations intrinsically humane and some of the ethical nuances and responsibilities feature clearly in the HOT Uganda project.

HOT Uganda participatory mapping process: how it works

HOT Uganda project design leverages the (perhaps unexpected) fact that in sub-Saharan Africa, mobile network is prolific. Almost every community in even the remotest locations in Uganda has – alongside the basic tools of existence (hoes, jerrycans, cooking pots) – a smartphone. Connectivity is everything and refugee settlements are clearly commercially viable for telecom installations, as evidenced by the plentitude of ‘booster masts’. Every dusty village has a highly effective informal and dynamic mobile-based economy. Wooden shacks are hand-painted with ‘MTN’ and ‘Airtel’ signs. ‘Mobile Money’ can be exchanged for airtime and mobile data can be exchanged for cash.

Community recruitment

The Uganda OSM project was built on a familiar model: training teams arrive in an open community environment, such as a marketplace or commercial centre, identify the busiest and best-equipped smartphone servicing trader and ask the (inevitably well-connected) dealer/repairer to round up 20 smartphone-literate acquaintances. But they must each have a smartphone. Smartphone possession encodes a number of characteristics about an individual: they can communicate, they are aspirational, technically engaged, sociable, and often represent an informal community group around the shared village computer.

Practice survey forms will have been written in a spreadsheet and uploaded using the simple coding of XForms to an online open-source server, to be deployed as app-readable page-swipe surveys (converted to Extensible Markup Language (XML) format at the server). Each person who has assembled then enters a race to first clear space on their phone, then download the necessary OSM-compatible apps (OSMAnd and ODKCollect). A survey is the next item to be downloaded and some mock interviewing of local people ensues. At each stage, eventual surveyors are identified on the basis of their aptitude. This shortlist of candidates is eventually selected to be the mapping group.

Uganda-specific practice

Uganda is a low- to middle-income country and decentralising leadership to local levels and promoting community self-sufficiency is a priority for the response coordination. The training of local administration in Uganda is coordinated with local community leaders. The HOT model implements training for all comers (including district officials) and is conducted in community/administrative centres. ‘Sensitisation’ campaigns with the local district and council authorities sound formal, but the benefits of making such comprehensive local data so openly available are always very warmly welcomed. Through this collaboration, community mapping can be legitimised within all-important local ‘community entry’ mechanisms. HOT inclusively adopts compatible tools like Microsoft Excel and ODK/Kobo to integrate (in fact, local government is informally run using tools such as WhatsApp). HOT also tries to engage stakeholder NGOs and host communities in collaboration, to allow both individuals and groups to create maps.

Part-time and occasional staff from local district administration often get deployed by HOT as surveyors, so that the experience can help them in their own institution. They will take new skills back to their work colleagues, as globally supported OSM members. Refugee and local hosting communities commonly work together, mapping the difference in services inside and outside of refugee-setting geographies. This common training environment also places NGO staff (often decision-makers) next to the beneficiaries they serve, and all participants are learning by doing (see Figure 9.4).

The socio-economic exclusivity of smartphone ownership is of utmost concern. There is a strong lending culture in Uganda (and many sub-Saharan settings), and it has been found that, invariably, people will find a way to be connected. Surveyors often turn up with a bad smartphone, but with good skills, and these people are subsidised by ‘project-loan’ smartphones.⁹ Technological inclusivity is acknowledged (and tackled) at the heart of HOT interventional planning. Depending on project-loan phone availability, field-budget funds will be diverted to enable device sharing; ‘earn-a-device’ schemes (per data output) have been initiated and analogue OSM tools will also get used, such as field papers. The need for subsidised devices continues, and solutions to the issue of digital access and inclusivity are constantly found on a location-by-location basis.

Classroom training

The first morning of a ‘capacitation’ (training) event is usually chaotic. There are generator breakdowns, technical glitches, computer viruses and cracked screens. Almost every training is chronically oversubscribed, but in the scrummage to learn, it is preferred to train as many as possible in OSM, despite the limited number of selections who will graduate as full-term surveyors.

The first exercise after assembling and registration is the presentation of the blank map, containing at this point only satellite-traced shapes, from remote-mapping sessions (in London, New York, Heidelberg, for example). This image can already be seen online (in Maps.Me and OSMAnd smartphone apps), but is printed out at this point. Participants are encouraged to write their name on a post-it note and stick it in the vicinity of their dwelling (see Figure 9.5). This is a good way to assess and introduce map literacy, but also gives an idea of where people come from and is used for local contact when later moving field surveying into that area.

This introduction emphasises how GIS information as OSM enables the delivery of humanitarian assistance in the form of engineering, medical and cultural intervention in areas generally considered inaccessible and precarious. There is a lot to be overwhelmed by on this first day. Not least is the task of bringing people through the digital revolution in one day. Clarifying the usefulness of OSM for the longer-term advantage for business development, personal enterprise and other socio-economic projects is also a challenge. All this with intermittent mains power, usually from an unreliable local market generator on an extension, involving bare wires wrapped in carrier bags for insulating tape. So much is inevitably improvised and the venue is usually a shack, tent or hall. Participants often have not used computers before and in some cases cannot speak a language common to one of the HOT trainers (however, surveys are translated into local languages and OSM itself exists in many international languages).

In the next stage, all trainees are signed up to OSM, using phone-tethered laptops (hard-wired infrastructure is always fragile in sub-Saharan Africa and wired telephones are scarcely used, 4G being more reliable than Wi-Fi service).

OSM is the online platform. Signing up, as with Facebook, gives the candidate sudden (if not yet conceived) access to OSM membership. This membership represents access to a huge online learning resource, with several professional-level curricula in modern GIS, all lodged on the ‘OSM wiki’ (Wikipedia, 2019c). More importantly, the new OSM member now has a potentially direct link to every OSM contributor – the expertise, skills, influence and assistance of hundreds of thousands of like-minded people, on essentially cash-neutral terms.¹⁰

WhatsApp groups

At the end of each training, a WhatsApp follow-up group is established, by which ongoing support can be given to the trainees and further training opportunities can be established and circulated. This is much more than a social group, however. For those selected as surveyors in this training, a second WhatsApp group will be formed. In a similar model to how local administration uses WhatsApp in Uganda, this will be the coordination tool used for future field-mapping. Outlying surveyors use WhatsApp to send map pins of their position (using OSM for Android/OSMAnd), important feedback relating to marginalised communities and also to transfer digital documents like their GPX (GPS Exchange Format)¹¹ tracks (collected in the OSMAnd app as they move). These will later be uploaded to OSM as roads/paths (sometimes roads are obscured from mapathon-goers by trees/clouds in satellite imagery).

Field training

This second day is an intensive day of field communication technique training with the best candidates selected from the day before. Real pre-prepared UNHCR/HOT surveys get workshopped and adjusted according to the area-specific services and hazards feedback gathered during the previous day’s training. Surveyors are split into groups of three – an interviewer, an interviewee and an evaluator. The groups do a round of rehearsal in each of these roles and then a feedback session is held, where comments are shared about the sessions and also about the forms that they will now depend upon (and suddenly have much more commentary on!). Certain questions are now understood to be irrelevant, inappropriate or badly worded (and get updated). One such adjustment was made when selected surveyors came back on day two and described survey changes that needed making: in Uganda, refugees are formally assigned blocks in refugee settlements. But in many instances, the question ‘What block is this?’ needed adjustment from ‘block’, to ‘tank’, because water tanks – installed albeit temporarily – had taken on community significance as meeting points and community identifiers. The survey was duly changed and a new global OSM convention was proposed. The focus of HOT Uganda’s work is partly on the collection of data, but also the communication of the terms attached to the universal spatial language of the map – the ‘legend’ – an arbitration of signifying systems by society; a dynamic connection of ‘text with technology’.

HOT trainers often learn techniques from these ever-resourceful locals, particularly smartphone tricks. Sophisticated techniques develop out of resource-poor necessity. Community surveying is an exercise in dramatic role play, and the most productive of all the sessions. It is also the first time that the importance of local semiotics of tagging and taxonomy is discussed.¹²

After some more rounds, surveyors go on foot into the surrounding area, accompanied by a trainer, and conduct surveys in the local village/block. When they return, their data (which in their survey app has been configured to automatically ‘send when finalised’), can now be seen on the online server (data collection and visualisation web page), with timestamps and locations attached. It can be viewed as a map, graph, list and many other options. This is when, for many of the trainees, the transparency and accountability of the process is revealed. They can claim that they asked a certain question in a certain place, but if the server places their geo-point somewhere else on the map, their colleagues can see otherwise. Initially, surveyors do not expect their work to be so visible and this can lead to a lot of hilarious interactions. Surveyors also manually report their surveys on the team WhatsApp group, as well as regularly checking in with a location pin to show where they are (live information is sometimes necessary for team security around border areas). As soon as the reports match the data seen by the field coordinators on the Kobo/ODK server, the remote mappers are paid for that day’s work, using another local technology, ‘Mobile Money’ (the mechanism for most personal banking in Uganda) (Wikipedia, 2018).

Motorcycle mapping

To optimise the speed at which this data can be collected, local forms of transport infrastructure are employed. At the end of this second day, surveyors are introduced to their new partner: the boda boda (motorcycle) rider. Motorcycle riders are selected, under supervision, by the surveyors, on the basis of geographic/local knowledge, language knowledge and professionalism. The two-person team then works together to strategise and navigate the region they are mapping. The robust security and logistics network of commercial boda boda fraternities is brought to the campaign by this partnership. With motorcycle mapping, collaboration takes place as occupational engagement between all participants, while the boda boda riders themselves are encouraged to learn and participate in the OSM process, bringing a number of fundamental assets, meanings and geographic practices to the table. Not least of these assets is the mobile infrastructure of the boda boda itself – particularly the battery, which is used to charge the team’s smartphone (see Figure 9.6).

(In)Formal field ontology: walking the walk

Northern Uganda fieldwork is punishing. It takes hard work and surveyors physically visit every single feature of every single village. It cannot be done using short cuts, or ‘telepresence’. The essential core of the project is a collaborative attitude and the ability to communicate. This will be what gets a field surveyor the data (and the acceptance) they need. Fieldworkers will be badly received in the rural communities if they cannot convey why HOT’s survey is different from the other hundreds of surveys more accessible communities regularly answer. To make data under the OSM orthodoxy useful to humanitarians, the data model remains continuously collaborative between HOT and UNHCR sector leads (Wiki OpenStreetMap, n.d.), in an attempt to maintain optimal alignment between OSM tags and UNHCR indicators. Local geospatial indicators apply to all of the humanitarian sectors. The surveyors, communities and interviewees all (ideally) continue to modify and collaborate on these as well. Five surveys are performed on each community, and then linked together by a sixth, introductory community profile survey that is carried out on arrival in each village or refugee block. Surveying is not an isolated event and happens in community discussions of how local priorities can be recorded.

The result of OSM mapping is a combination of both the official coordination data requirements and how the refugee community want to identify their needs for themselves. So informal interventions are also mapped; traditional healers get mapped and tagged as ‘health features’ alongside large NGO hospitals; dug-out wells in dry riverbeds (WaSH) are mapped and detailed (see Figure 9.7) alongside formal bore-hole installations (see Figure 9.8).

Mapping all features that are tagged as ‘non-functioning’ precisely identifies service deficiencies, but integration of WaSH and education cross-sector information can be visualised in walking distances to water points and schools to predict school truancy (children are traditionally the household water collectors). Maps that can show flooding wells lying close to flooding latrines have a very real impact on cholera outbreak control. A well-known cross-sector analysis of data in context is on unlit latrines (UNHCR, 2017a). Women (a large proportion of the population¹³) report sexual violence when using services at night and these hotspots can now be geographically pinpointed (REACH/UNHCR, 2018). Resources such as trauma care (health) and lighting (physical planning) can be allocated to latrines/wells (WaSH), per location (protection).

These combinations of data are endlessly useful (see Figure 9.9) and the potential basis for future disaster preparedness (an Ebola outbreak, for instance) is provisioned for. This kind of data is received with open arms by local field offices, and at this point, the project shows an openly compatible geospatial platform connecting historic geospatial records and new community witnessed data with the outside world: connecting the ‘specific’ to the ‘universal’).

Mission accountability and financial transparency

Field-mapping campaigns can continue for weeks and months, with diminishing need for on-site supervision, as new leaders emerge in the teams. HOT have scaled projects into new areas by transferring increasingly skilled community members to lead. Many of the refugee and host participants come from violent and mendicant cultural settings where, after generations of (cross-border) war, the value of long-term planning let alone job security is seldom experienced. Surveyors can be put off by the exacting nature of the role, but a very satisfying work ethic often starts to evolve as daily wage payments stay transparently regular and ‘their’ map gets visually filled in and becomes visible on their smartphone app. As the surveyors graduate into fieldwork, they also start learning data cleaning from coordinators.

Accountability is conceived in other directions too. For humanitarians, it is fundamentally a transparent, cost-neutral project by which donors can engage and collaborate with their field counterparts, giving time rather than money to support the production of commonly owned visualisations from satellite and field data. Remote mappers as ‘donors’ in far-off mapathons see features that were simple shapes, lines and points getting names and attributes attached and indicating ground-truths.

This hyper-local accuracy demands participation. It involves surveyors and their communities in making decisions. Data collected in the field by HOT in Uganda enjoy strict data cleaning and total control at source. OSM can only share geographical and technical information about features themselves. However, certain data such as population or numbers (and locations) of disease cases may comprise vital narratives for responders. Surveys are therefore split by column of data type in a spreadsheet and human and demographic data are carefully anonymised and secured with different permissions. This separation effectively secures the system. It is how HOT are able, in conjunction with local police, to securely manage victim self-reporting of female genital mutilation in Tanzania.¹⁴

The argument for lo-fi tech in humanitarian interventions

HOT has no direct competition in tackling the many geospatial issues faced by the UNHCR (digital inclusion, community engagement, cultural integration, accountability). But is OSM digital humanitarianism? Can its post-digital capacities offer insights into disentangling the problems of humanitarian digital telepresence? Or is HOT’s brand of crowdsourced innovation just another contributor to the problem of innovation overload? Certainly, the revolutionary nature of device-based participation is not a catch-all humanitarian solution.

Glasze and Perkins (2015, p. 144) argue that ‘researchers might also deploy mixed and ethnographic approaches, in order to learn more about particular moments of mapping practice’. HOT’s data are technically accurate (up to 0.6 metres current smartphone accuracy and locally triangulated survey checks), but it is not 100 per cent ‘true’. Nor does it claim to be. The process of collaborative/collective mapping is a strong ‘legible’ example of how technical solutionism can form an important consensually agreed situated knowledge between two or more cultures. Open data is ‘in process’: it is iterative and reiterative. Field mapping involves constantly changing versions of the map areas to be printed, used for navigation, adjusted and updated through digital upload again; then redownloaded and used for planning and navigating once more.

OSM is not made of software short cuts. In fact, it explicitly relates to the physical. It is by no means intrinsically digital,¹⁵ but a some-time analogue, post-technological interpretation of everyday surroundings, a creation of human connections within communities in need. Without physical presence, HOT could never hope to have the impact it has. On the ground in Uganda, HOT’s mantra is ‘Local people, local tech, just add knowledge’ (HOT, 2018). OSM has important analogue tools that use post-digital locally inclusive technologies such as pen and paper. ‘Field papers’ – A4 printouts of sections of OSM maps – print QR (quick response) code geolocators on to each of the map sheets. Once hand-drawn and written information is noted on the paper in the field, a photo or scan of it can be digitally geolocated by OSM editor software (e.g. Java OpenStreetMap Editor – JOSM). Editing can then be done between the paper image and the satellite image and merged as edits to OSM.

Cross-fertilising qualitative and quantitative data

There are various ways to check qualitative data and the ideal is to combine it not only with quantitative data, but with diverse approaches and opinions. Highly efficient dispersed consensus is reached by using the community themselves, having them work together and cross-check each other’s work. One key component of the Uganda community mapping project is the process of returning paper maps to settlement zones and sub-counties where the data originated, for review, further input, cultural ownership and continued participation.

To mitigate for misinformation (e.g. indiscriminate denial of any services in a community, to leverage more NGO finances), debate is triangulated by the surveyor in the public outdoor space of the village/settlement. Data are then cross-checked later through other local sources, and callbacks are conducted (using the contact number collected with the geo-point). This makes for a triangulated quality-checking process. In OSM, data are there to be contested, interacted with, then corrected; then updated again, then contested, ad infinitum. The terms, or ‘taxonomies’, are proposed for and agreed by consensual majority.

Individuals will sometimes give misinformation that they believe is expected, or for personal gain. HOT’s ability to comprehensively map every single feature on the ground in every community has meant that real numbers and percentages can be cross-referenced with reported attitudes and perceptions of these realities projected from quantitative sample surveys. OSM-engaged NGOs are asked for their input as an organisation and often share qualitative data with the team.

Extrapolations from qualitative data are extremely useful for all aspects of humanitarian decision-making. The mass smartphone-enabled quantitative data allows not only evaluation of this, but radically changes the spectrum of possible analytics. Very basic data analysis can reveal extremely accurate numbers of people affected in exact locations, by assessing qualitative data on population per household and remote-mapped shelter imagery. In an ideal world, both qualitative and quantitative formats of data are used in tandem like this, for a rich and informed picture. Data are useless without analysis, and again, the truly useful information is to be found in the contest between the two types of data. Questions arise such as: ‘Why are these data different?’ and ‘What issues are we missing here?’ Qualitative data may express or contradict a narrative clearly visible in physical quantities (e.g. of patients near hospitals, teachers per schoolchild, refugees per settlement), and quantitative data should always be analysed in terms of what qualities they convey.

In all instances of humanitarian crisis, fieldworkers unanimously agree that any data is better than no data at all, and the ‘rapid-cycle mapping’ (Johnson et al., 2010) work of projects such as HOT/Missing Maps comprises an invaluable resource for most emergency data needs – even when incomplete. In practical terms, open data arguably benefits from its own self-professed fragility and this contest leaves important space for empathy and interpretation.

Principles of contest and consent: choice and community protection

The evolutionary nature of the ongoing and dynamic map is a major part of the HOT message in Uganda. Surveys collaborated for the community (by the community) adopt local terms to identify hyper-local assets. Unlike authoritative data, it is mobile, interactive and free from the requirement to claim fixed accuracy. That it claims no reassurances makes for the argument of a safer system in terms of credibility, trustworthiness and expertise.¹⁶ OSM demands interaction, and is available on smartphone apps (Maps.ME, OSMAnd), by which it can be endlessly updated. Surveyors see new uploads as soon as the latest updates appear online. The reciprocal relation between local and global technologies and local and global communities is key. It is processual and internationally collaborative. It is a conversation.

Whether or not the participatory nature of OSM positively recolonises the map and its users, it does at least engage the users in a debate of its own maps and the proverbial ‘ink with which they were written’. The personal ‘epistemic disobedience’¹⁷ afforded to each human participation to OSM is a highly important component of protection and representation. The participant becomes invested. People can delete their data on OSM. Hyper-local accuracy demands participation and it involves surveyors and their communities in making decisions.

This physical human choice is vital to mediate the digital machine at either end. The imperative to intervene has been encountered by HOT in other settings, where under changing political landscapes, undocumented immigrant data can fall into different hands over time. OSM data can be dangerous too. For example, in an informal settlement in Zimbabwe, local knowledge of 500,000 unofficial residents was mapped, but accuracy was restricted to only neighbourhood-level resolution. This vulnerable undocumented community needed representing, but although OSM was technically capable of identifying individuals house by house, mapping them would potentially expose them as well as represent them. The last few metres of tracing individuals was done by spoken word, ultimately maintaining anonymity.¹⁸

The capacity for participator feedback is key to responsive and responsible mapping. What cannot be forgotten is the physicality of communities and people in the field, and the difficult but rewarding work of the surveyors in translating this reality. Dayan, now HOT field team leader for West Nile, reported diplomatic discussions in which refugees, living in often bullet-strewn border communities, fear being surveyed, ‘since most of them are traumatised’. He explains that the nature of the mapping work finds that ‘many [of the rural villagers] are political or military’. He went on to report:

Textuality, addressing, semiotic coding

Paglen (2009, p. 1) writes that: ‘In a nutshell, the production of space says that humans create the world around them and that humans are, in turn, created by the world around them.’ As ‘poster-child fetishism’ or ‘disaster tourism’, static snapshots can quickly come to misrepresent. The concept of ‘address’ as verbal text quickly comes to have a lived (interventional) impact in humanitarian information management. So it is imperative that the interactive nature of digital mapping can represent – but also protect – those it serves. Open Data Kit (ODK) digital survey forms are a way of linking words to objects in a coded material way. OSM mapping practice seems to show that chosen words, descriptions and explanations are just as important as technological empowerment.

As mentioned above, communities under survey sometimes express answers that clearly reflect how they would like to be seen and perhaps not how they actually appear to an outsider. OSM allows for plural expressions of reality on the ground and often finds communities preferring more unexpected indicators. An example is local insistence on anglicised – not tribal – spellings of village names. The reason for this pedantry is the desire for information to be known and trusted in the right way by decision-makers, which may have a visceral impact in life-threatening situations of emergency administration.

Reality, representation and counter-mapping

In The Freedom of the Migrant, Wilem Flusser (2003, p. 86) writes that ‘the creation of new information depends on the synthesis of prior information. Such a synthesis consists in an exchange of information … One can therefore speak of creation as a dialogic process, in which either an internal or an external dialogue takes place’. Texts such as Mythologies (Barthes, 1973) questioned the existence of a singular extrinsic reality, and this (and other) studies of semiotics investigated how ordered relations (codes) of significance create their own meaning. The deconstruction of this idea that humans are subject to only one (dominant) reading of meaning created, in both philosophical and practical terms, space to effectively reinvent meaning. The social constructivist project of OSM enacts this conception through address, creating a language of action and meaning within tagging conventions. OSM discourse tries to accommodate plural ontologies of place, but convey a unified meaning (even if that is characterised by contest). Interpretation happens in the construction of relevant survey questions, but also in the collection of those data. A transposition of ‘the virtual’ to ‘the real’. As Barnes and Duncan (1992, p. 6) put it: ‘A landscape possesses a similar objective fixity to that of a written text … [becoming] detached from the intentions of its original authors … various readings of landscapes matter more than any authorial intentions … [are] constitutive of reality, rather than mimicking it.’

Through the creation of the ‘legend’²⁰ of the OSM map, we can infer that participants are subjects, auto-ethnographers and authors, simultaneously. OSM sets up the map landscape as an ‘interrogative text’, ‘disrupt[ing] the unity of the reader by discouraging identification with a unified subject of the enunciation’ (Belsey, 2003, p. 91). Meaning-making in OSM presents a condition of contest and consent, but it is vital that this interrogative feedback loop be maintained.

Satellite gaze: power, smoke and mirrors

In the years leading up to the launch of GPS, thinkers like Foucault ([1966] 1970) popularised disruptive readings of spatial intersubjectivity, rewriting object–subject relations in cultural sites of ‘gaze’ such as Diego Velázquez’s painting Las Meninas, an interrogation of dynamic politics of intersubjective gaze in art and contemporary cultural theories. Mass production of GPS made available a technology that could similarly politicise the shifting gaze of international intelligence. So substantial has been this ‘prospect-shift’ that eventually the preoccupation with ‘who is looking at whom’ arguably diverts even heads of state themselves from linear notions of truth or news and they can become associated with ever-receding versions of populist reality.

Byung-Chul Han (2017, p. 1) writes that, ‘Today, we do not deem ourselves subjugated subjects, but rather projects’ (original emphasis). Expansions of Jacques Lacan’s ‘real’ into ‘symbolic real’ (Žižek, 1999, pp. 222, 276) debate how media constructions of consensual and interactive reality influence contemporary understandings of ‘the self’. Certainly interactive media overtly exploit ambiguities in ‘mapped truth’ (e.g. the Pokémon GO computer game using OSM), thereby positing a version of reality altogether different from the version that traditionally presides.²¹ Consent lies in the ‘common land’ of a space, but in a world in which terms like ‘augmented reality’, the ‘real’ and ‘the hyper-real’ are now in slippage, critics like Žižek (1999) contend that interaction, or a Foucauldian address process, secures credibility of truth: space has become a process, rather than an object.

Studies of human production of space have traditionally associated the perception of landscape texts with social and political manipulation. Jay Appleton’s (1975) analysis of landscape poetics describes how estate landowners during periods of imperial growth in Britain demonstratively implied dominance by installing elevated buildings and watchtowers on landscaped grounds. This spatial aesthetic of cultural dominance over estate was coined ‘prospect theory’. It features in subsequent writings on territory and colonial gaze (Fitter, 1995; Turner, 1981). Prospect theory was interestingly extended by Appleton (1975) to the concept of ‘indirect prospect symbols’: viewpoints overlooking (imaginary) territories at the peripheries of large country estates. According to Appleton (1975, pp. 80–1), these ‘symbolically invite the speculation that they command a further field of vision’. Through this extension, not only do you govern all you can see, but by implication, all that can be seen from your outlying viewing positions.

So how does this apply to satellites and their cameras? Discussion becomes interesting when ‘selfie’ culture is applied to the placement of the satellites as a technological entity and the semiotics of ‘indirect prospect’ take on an extra dimension. According to Senft and Baym (2015, p. 1589), the selfie is a

As a GIS, the OSM community effectively sees itself through its own satellite cameras. Humans, in selfies, look at imagery – effectively of themselves – but through the lens of their own machine. And that gaze becomes modified by the act of looking. The communication of prospect here, too, becomes reciprocal between subjects and objects, and exploits that which (traditionally) has exploited it. In Bidibidi, the example of community agency was documented in the BBC series Equator from the Air.²² The feature concerned itself extensively with the reciprocal relation between recording and creating the peri-urban order that is coming to signify ‘Bidibidi the city’.²³

Subject, object, fetish and factish

The language of address and self- address in the theatre of crisis has long made reality itself a contest, with the process of its creation becoming increasingly inclusive. According to Scott Blinder (2012, p. 4), ‘public opinion toward immigration is directed toward “pictures in our heads” of immigrants rather than immigration per se’. If the communities in Uganda own each survey data set, images from the field may, under the OSM rubric, reappropriate ‘disaster tourism’. The representation of the subject in humanitarian intervention is something in which persons of concern take a keen interest. Selfie culture has become an intrinsic part of the media production–induction–consumption loop, but what may seem a superficial product of northern hemisphere narcissism is, not only ubiquitous but culturally contingent.²⁴ Loaned smartphones return from the field full of surveyor selfies.

The underlying message of the OSM processes in Uganda is one of reclaiming the map, with situated knowledge as a fundamental practice. ‘Language is the medium through which a hierarchical structure of power is perpetuated, and the medium through which conceptions of “truth”, “order”, and “reality” become established … such power is rejected in the emergence of an effective post-colonial voice’ (Ashcroft et al., 1989, p. 7). This voice finds itself decolonial in OSM (Mignolo, 2009). The data of the community belongs to – and in – the communities that it represents. With an open market on digital engagement itself, the responsibility to engage can now tackle real-time issues of misrepresentation, engaging with ‘voluntourism’ – and even cultural appropriation²⁵ – in the process.

OSM advocates could argue that, possibly for the first time, the subjects of disaster mapping may hold full control (and consent) over the artefacts of their own data. Bruno Latour’s (1999, pp. 1–23) exhilarating discussion of cultural appropriations of myth in ‘the slight surprise of action’ opens up this meaning-making and ownership process a little further. It contemplates fetishisations as sites where collisions of object and of place seem inevitable. Latour discusses ‘fetish’ becoming ‘factish’ in the contest of cultural appropriation and iconoclasm. Ultimately, he argues that the object gains meaning only once that very meaning is challenged by the iconoclast. Paradoxically then, the iconoclast enables the meaning s/he objects to: a cyclical generative and degenerative process.

Latour uses the example of the Nepalese legend of the Jaggernath and the magic stone, in which the teacher (Jaggernath), seeks to emancipate his community from subjugation in the worship of an untouchable stone (the saligram). The protagonist breaks the taboo of the saligram²⁶ by picking it up and forcing the awestruck community members to touch it too. The action provokes a revulsion not towards the controlling myth of the stone (which remains, if anything, more untouchable and sacred), but towards the attempted emancipation of the myth-busting process itself. Most of all, the essay describes the dehumanising of the iconoclastic protagonist, Jaggernath. And so, identifying components of meaning-making here (or to refer to Tim Ingolds (2011), the ‘ink with which it is written’) may not be what the community wants. Equally, post-structuralist camps might argue that society depends on consensual mythology to maintain social cohesion, and that there exists an intrinsic dependence upon ‘legend’ for meaning as a whole.

An iconographic parallel may be found in the representational politics of an installation in the Namibian desert, in which a solar-powered sculpture, with an MP3 player embedded, plays a sound reproduction of Toto’s 1982 hit ‘Africa’ on a loop: ‘The artist set up an installation called Toto Forever, made up of six speakers attached to a blue MP3 player – whose only song is Africa, set to play on an infinite loop – all standing atop white rectangular blocks set up in the sand. The installation, Siedentopf writes on his website, runs on solar batteries “to keep Toto going for all eternity”’ (Aratani, 2019).

The song ‘Africa’ by the band Toto is a cultural appropriation of some distinction, which seemed to smooth out the squalor, injustice and discomfort of a ‘real’ Africa for a 1980s pop music audience. The song is noted playing in African bars by some who spent enough time there to remember how this romanticism of Africa from afar seemed incongruous with the realities of Ethiopian famine, apartheid dissolution and the West African civil wars. However, played in bars it is. And owned and loved by Africans, too. There is a ‘reterritorialised’ capacity to this, one could argue, in which its semantic script has been rewritten. And like the installation in the desert, the iconography has come to define the continent that, contested by all who engage, wilfully claims its right to a decolonial description or ‘tag’. And this seems disobedient. Ironically, the main visual feature in Toto’s pop video from the same year is – wait for it – a map.

Mobility, resilience and united statelessness

A conception of ‘home’ as a ‘place’ (rather than a ‘process’) implies threat to many in the world who do not self-identify as refugees. But mobility of meaning does not seem here to challenge social cohesion, but rather make it stronger, through a debate of reciprocated relativist self-professed fragility. Stability in this territory comes from aggregated risk, resilience from connectivity and impact from participation.

Contemporary debate on the grey area between enforced and voluntary migration ‘must account for the multi-causality of population movement’ (Parater et al., 2019). Although traditional assumptions link displacement with disadvantage, global societies increasingly question the idea of home as a place, citing various reasons for embracing geographic fluidity. Questions around choice in itinerance are found across cultures, and these cultures are beginning to demonstrate connections. Digital nomadism is prized by some as an uber-class luxury in non-disaster settings, and in fact the ethos of this nomadic practice favours similar values of creativity, improvisation and community self-governance, which seem missing from the contemporary mainstream. OSM potentially connects these groups across cultures.

Discussion in this chapter has been concerned with mobility: mobile meaning, mobile infrastructure and mobile networks (in more than one sense). The lack of formal infrastructure in many sub-Saharan African settings is rapidly emerging as something of an advantage in a sociopolitical environment of mobile lifestyle (and even mobile meaning). Resilience derives from adaptability. Some might say that the historical fragility in many parts of Africa renders populations adaptable, immune to reliance on hard-wired infrastructure (e.g. copper telephone lines) so counted on by the global North. Mobile networks (commercially viable on a free market) make itinerance relatively dependable. Although settlement environments are deemed precarious, reliability is something that is immediately visible by line of sight (booster masts are prolifically installed). Here, network is physically present, not a blind function of unseen bureaucracy. And here, too, commercial interests can be legitimately hacked for social advantage.

Conclusions

Perhaps OSM collaborators create terms that challenge inevitable subjectivity, where the very fact of intersubjective expression can empower a liberated language of action. The terms afforded by OSM are owned by the local communities and their spokespeople. In Sierra Leone, motorcycle mapping has revealed villages self-identifying their particular locations as under one chiefdom even though the territory itself is spatially isolated from the main geographic body of that chiefdom: an island of local geopolitical significance.

Nobody – and yet everybody – owns OSM. The auto-ethnographic distributed quality of OSM is, perhaps, a coping mechanism in a world where truth is increasingly fragile. One of the engagements that has helped OSM to be accepted in Uganda is ongoing HOT collaboration with Uganda’s National Bureau of Statistics (UBOS) (Allan, 2018). Serious efforts are now afoot to undertake the next national census using OSM tools. If this is implemented, it would make Uganda a global pioneer in the adoption of these accurate, cost-effective and sustainable methodologies (and, in a cartographic sense, the most plurally inclusive nation on earth). It is clear that Uganda has a need for a robust, self-reliant and sustainable infra-structural system using OSM, which can have shared and minimised cost. This need is demonstrated through support from the World Bank and various UN agencies, but intervention methods used in Uganda need to merge with – rather than disrupt – existing local administrative systems. Responsibility needs devolving across citizens and sectors alike. For this, the accountability and granularity of OSM needs to be seen as enabling, rather than threatening.

OSM has been called a ‘do-ocracy, meritocracy, technocracy, and bureaucracy … a radical change that is significantly different from other digital mapping projects’ (Glasze and Perkins, 2015, p. 153), and it does seem a viable model for practical social constructivism. Solutions to crises are perhaps to be found in more human engagement with technology, but traditional social mechanisms seem able to reboot too, in this emerging post-digital mode. OSM’s dealings with fragility firmly locate it in the realm of the self-organising systems which Connolly (2013, p. 119) identifies, of ‘plural assemblage … composed of those sharing affinities of spirituality across differences of creed, class, gender, sexual orientation, and ethnicity’.

Innovative technical solutions (hacktivism, liberation technology) increasingly find post-digital recognition in self-governing communities like OSM, thereby making the ‘radical desire to change the world’²⁷ less and less restricted to the economically privileged. A future where borders, boundaries and terms like ‘alien’ become redundant envisages global mobility, post-capitalist inclusivity and peer-to-peer pro-poor social cohesion becoming part of everyday existence. In the light of actor-network theories of social cohesion, those preoccupied with containment of territory and displacement may learn from the more resilient models of cooperative and adaptive community interventions to define mobility in its gathering global momentum as an inclusive expression of defiant and united statelessness.

References

Allan, R. (2018) ‘Uganda Bureau of Statistics engages in open mapping for resilience with the OpenStreetMap community’, 1 Oct., https://opendri.org/uganda-open-mapping-for-resilience (accessed 3 July 2019).

Appleton, J. (1975) Experience of Landscape (Hoboken, NJ: Wiley & Sons).

Aratani, L. (2019) ‘Toto forever: Africa to play “for all eternity” in Namib desert’, Guardian, 15 Jan., https://www.theguardian.com/world/2019/jan/15/toto-africa-desert-installation-play-for-all-eternity (accessed 15 January 2020).

Ashcroft, B., G. Griffiths and H. Tiffin (1989) The Empire Writes Back: Theory and Practice in Post-Colonial Literatures (London: Routledge).

Balcik, B., B.M. Beamon and K. Smilowitz (2008) ‘Last mile distribution in humanitarian relief’, Journal of Intelligent Transportation Systems, 12 (2): 51–63.

Barnes, T. and J. Duncan (eds.) (1992) Writings Worlds: Discourse, Text and Metaphor in the Representation of Landscape (London: Routledge).

Barthes, R. (1973) Mythologies, trans. A. Lavers (New York: Hill & Wang).

Belsey, C. (2003) Critical Practice (London: Routledge).

Blinder, S. (2015) ‘Imagining immigration: the impact of different meanings of “immigrants” in public opinion and policy debates in Britain’, Political Studies, 63 (1): 80–100.

Brandt, R. (2014) ‘Google divulges numbers at I/O: 20 billion texts, 93 million selfies and more’, Silicon Valley Business Journal, 25 June, https://www.bizjournals.com/sanjose/news/2014/06/25/google-divulges-numbers-at-i-o-20-billion-texts-93.html (accessed 28 March 2020).

Cassano, J. (2014) ‘Inside the crowdsources map project that is helping contain the Ebola epidemic’, 22 Oct., https://www.fastcompany.com/3037350/inside-the-crowdsourced-map-project-that-is-helping-contain-the-ebola-epidemic (accessed 2 March 2019).

Connolly, W.E. (2013) The Fragility of Things: Self-Organizing Processes, Neoliberal Fantasies, and Democratic Activism (Durham, NC: Duke University Press).

Cosgrove, D. and S. Daniels (1989) The Iconography of Landscape (Cambridge: Cambridge University Press).

D’Angelo, J. and B. Ranalli (2019) ‘The dark side of sunlight: how transparency helps lobbyists and hurts the public’, Foreign Affairs Magazine, 98 (2), https://www.foreignaffairs.com/articles/united-states/2019-04-16/dark-side-sunlight (accessed 23 June 2019).

Dirk, H.R., J. Birckhead, D.G. Green and J. Atkinson (1996) ‘The electronic colonization of the Pacific’, Computer Mediated Communications Magazine, 3 (2): 1–9.

Fitter, C. (1995) ‘Toward a theory of landscape’, in C. Fitter, Poetry, Space, Landscape: Towards a New Theory (Cambridge: Cambridge University Press), pp. 1–24.

Flusser, W. (2003) The Freedom of the Migrant (Urbana, IL: University of Illinois Press).

Foucault, M. [1966] (1970) The Order of Things (London: Pan Books).

Givoni, M. (2011) ‘Humanitarian governance and ethical cultivation: Médecins sans frontières and the advent of the expert-witness’, Millennium, 40 (1): 43–63.

— (2016) ‘Between micro mappers and missing maps: digital humanitarianism and the politics of material participation in disaster response’, Environment and Planning D: Society and Space, 34 (6): 1025–43, https://doi.org/10.1177/0263775816652899.

Glasze, G. and C. Perkins (2015) ‘Social and political dimensions of the OpenStreetMap Project: towards a critical geographical research agenda’, in J.J. Arsanjani, A. Zipf, P. Mooney and M. Helbich (eds.), OpenStreetMap in GIScience (Lecture Notes in Geoinformation and Cartography) (New York: Springer), pp. 143–66.

Han, B. (2017) Psychopolitics: Neoliberalism and New Technologies of Power (New York: Verso).

Harvey, D. (2009) Social Justice and the City (Athens, GA: University of Georgia Press).

Humanitarian OpenStreetMap Team (HOT) (2018) Participatory Mapping Toolkit: A Guide for Refugee Context, https://www.hotosm.org/downloads/Toolkit-for-Participatory-Mapping.pdf (accessed 17 March 2020).

Ingold, T. (2011) Being Alive: Essays on Movement, Knowledge and Description (Abingdon: Routledge).

Johnson, J., J. Crowley and S.E. O’Reilly (2010) ‘Where: 2:0 – Haiti: crisis mapping the earthquake’, YouTube, 2 April, https://www.youtube.com/watch?v=fJvR84UX5RI (accessed 25 May 2019).

Latour, B. (1999). ‘Slight surprise of action’, in B. Latour (ed.), Pandora’s Hope: Essays on the Reality of Science Studies (Cambridge, MA: Harvard University Press).

Lefebvre, H. (1968) Le droit à la ville: suivi de espace et politque (Paris: Anthropos).

Mignolo, W.D. (2009) ‘Epistemic disobedience, independent thought and decolonial freedom’, Theory, Culture & Society, 26 (7–8): 1–23.

Muttaqien, B.I., F.O. Ostermann and R.L.G. Lemmens (2018) ‘Modeling aggregated expertise of user contributions to assess the credibility of OpenStreetMap features’, Transactions in GIS, 22 (3): 823–41, https://www.researchgate.net/publication/327098680_Modeling_aggregated_expertise_of_user_contributions_to_assess_the_credibility_of_OpenStreetMap_features (accessed 4 June 2019).

Nyabola, N. (2016) Digital Democracy, Analogue Politics: How the Internet Era is Transforming Politics in Kenya (London: Zed Books).

Okiror, S. (2018) ‘“They exaggerated figures”: Ugandan aid officials suspended over alleged fraud’, Guardian, 8 Feb., https://www.theguardian.com/global-development/2018/feb/08/they-exaggerated-figures-uganda-aid-officials-suspended-over-alleged-fraud (accessed 2 March 2019).

Paglen, T. (2009) ‘Experimental geography: from cultural production to the production of space’, Brooklyn Rail, 6 March, https://brooklynrail.org (accessed 2 March 2019).

Parater, L., A. Neimad and A. Christiano (2019) ‘Communicating the complexity of displacement in a changing climate’ (UNHCR Innovation Service), Medium, 27 March, https://medium.com/bending-the-arc/communicating-the-complexity-of-displacement-in-a-changing-climate-9c1ac0bf5e42 (accessed 28 June 2019).

REACH/UNHCR (2018) ‘Uganda joint multi-sector needs assessment’, https://data2.unhcr.org/en/documents/details/65982 (accessed 25 May 2019).

Redfield, P. (2013) Life in Crisis: The Ethical Journey of Doctors Without Borders (Berkeley: University of California Press).

Said, E. (1978) Orientalism (New York: Pantheon).

Senft, T. and N. Baym (2015) ‘What does the selfie say? Investigating a global phenomenon’, International Journal of Communication, 9: 1588–606, http://ijoc.org.

Turner, J. (1981) The Politics of Landscape: Rural Scenery and Society in English Poetry, 1630–1660 (Cambridge, MA: Harvard University Press).

United Nations High Commissioner for Refugees (UNHCR) (n.d.) ‘Uganda comprehensive refugee response portal’, https://ugandarefugees.org/en/coutry/uga (accessed 5 July 2019).

— (2017a) ‘How night-time street-lighting affects refugee communities: a population-based assessment of community lighting in Northern Uganda’s Rhino Camp refugee settlement’, Dec., https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2ahUKEwiBl-Hf4qDjAhVop1kKHd_PB6UQFjAAegQIARAC&url=https%3A%2F%2Fwww.unhcr.org%2F5b3cb5bb7.pdf&usg=AOvVaw1sowamWayKJAF2LEYnFG_u (accessed 5 July 2019).

— (2017b) ‘South Sudanese refugees in Uganda now exceed 1 million’, https://www.unhcr.org/en-us/news/stories/2017/8/59915f604/south-sudanese-refugees-uganda-exceed-1-million.html (accessed 3 July 2019).

UNHCR/Office of the Prime Minister (OPM) (2018) ‘OPM and UNHCR complete the countrywide biometric refugee verification exercise’, 30 Oct., https://reliefweb.int/sites/reliefweb.int/files/resources/66545.pdf (accessed 30 May 2019).

Wiki OpenStreetMap (n.d.) ‘Data model at OpenStreetMap Uganda’, https://wiki.openstreetmap.org/wiki/WikiProject_Uganda/Uganda_Crowdsourcing_Non-Camp_Refugee_Data (accessed 2 June 2019).

Wikipedia (2018) ‘African Mobile Money’, 27 Nov., https://en.wikipedia.org/wiki/African_MobileMoney (accessed 2 June 2019).

Wikipedia (2019a) ‘GPS Exchange Format’, 20 Feb., https://en.wikipedia.org/wiki/GPS_Exchange_Format (accessed 2 June 2019).

Wikipedia (2019b) ‘OpenStreetMap’, 20 Feb., https://en.wikipedia.org/wiki/OpenStreetMap (accessed 2 June 2019).

Wikipedia (2019c) ‘WikiProject Uganda’, 23 Jan., https://wiki.openstreetmap.org/wiki/WikiProject_Uganda (accessed 2 June 2019).

Wolbers, J. and K. Boersma (2013) ‘The common operational picture as collective sensemaking’, Journal of Contingencies and Crisis Management, 21 (4): 186–99.

Žižek, S. (1999) The Ticklish Subject: The Absent Centre of Political Ontology (London: Verso).