Software Ethnography: Notes from the field

This blog post was originally written to introduce the ‘software ethnography’ strand of the Disconnected Infrastructures project at our first advisory board meeting in early November. Jo Entwistle (read her first blog post here) and I are now a week into the initial fieldwork in Delhi (7-19 Dec 2017), at the offices of our partners, SafetiPin.

Making connections: Digital sociologies

It’s hard not to start from a personal reflection: the last time I worked extensively in India – 2002-2005 – I was part of a UNESCO programme called ICTPR (ICTs for Poverty Reduction),  which funded eight community media projects around South Asia, each focused on different ways of conceptualising the connections between information technologies (both new and old) and poverty reduction (Slater, 2013).

Many of the more techie questions we are asking in the Disconnected Infrastructures project simply could not be asked then: in most cases we never managed to get a reliable internet connection; often the computers couldn’t be powered; there were few phones and of course no smart phones; and – above all – very rudimentary understandings of the affordances of all this future kit, and how to place them in social and development processes. Fifteen years later our collective research is premised on being able to generate big data streams from smartphones in low-income areas of south India to servers in Delhi….

But there were at least two features of that older project (read the final report here) that seem very close to what we are doing now:

  1. At that time ICTs were themselves a disconnected infrastructure. Like smart, the development agency claim was that putting computers in development contexts would connect everyone and everything – people, resources, information society. In fact, computers appeared in community projects as alien imports that did not connect with anything else at all. To the contrary most development agencies and NGOs assumed that no proper communication systems even existed wherever they were working; that development sites could be treated as communicative tabula rasa, that there was nothing to connect ICTs to. Our job was to see how new information infrastructures could connect or integrate with existing ways and means of doing things, how to understand them as part of infrastructural ecologies of communication, transport, welfare, familial support, sanitation, housing and so on.
  2. ICTs were a material culture through which diverse people were coming to understand ‘society’ differently, and to model their personal and family strategies for dealing with an emergent new world. Just as with the idea of ‘smart’ and big data today, people engaged with the idea of ICTs as images of the future society they would have to deal with. My abiding memory is of very poor women spending painful proportions of their income on computer literacy courses for their sons and daughters, as ten years before they would have bought English classes or secretarial training – with very little knowledge of how computers might link to social progress. And even less idea of how different ideas of the social were baked into the technology.

In these two senses at least this project is not far off the last one: the key socio-technical questions concern:

  1. How can new technological systems – here software, ‘smart’ urbanism, new data flows – contribute to building liveable urban ecologies, by connecting in locally appropriate ways?
  2. How social representations, assumptions, images of the social, of how society works or will work, are hardwired into new technological arrangements, and what images of the social are projected onto these technologies?

Working with, and researching, Safetipin

(c) Safetipin 

SafetiPin have least two roles in the Disconnected Infrastructures project:

Firstly, SafetiPin are research partners: the organization has designed a complex information system for conducting and communicating urban ‘safety audits’, covering a growing roster of cities in the global South. The system is designed to source and code data on infrastructural and spatial parameters that impact on people’s feeling of safety in public spaces. This data is both represented on maps and aggregated into ‘safety scores’. It be can be used for lobbying municipalities and policy makers for infrastructure improvements; and to help publics – particularly but not exclusively women – to make more informed assessments of where or whether they will feel safe as they move about cities. At the endpoint of this information system is the SafetiPin smartphone app which GIS mapped data on individual and aggregate safety scores, as well as additional functions such as real time-tracking of friends and family, and a facility to choose the ‘safest’ route to a destination. Data is collected through crowdsourcing, through training local volunteers to carry out safety audits, and through coding photographic documentation of an area (smart phones are mounted on taxis/Ubers to photograph streets at 100m intervals; the images are then coded for the standard safety audit parameters and mapped along with the other data streams).

In this role, SafetiPin are research partners contributing data on infrastructure and perceptions of safety in our Keralan fieldsites, data that will sit alongside qualitative data generated through interviews and participant walkabouts on site.

This leads directly to the second part played by SafetiPin: they are the ‘object’ of our ‘software ethnography’. SafetiPin is a well-known and well-regarded project of bottom-up data sourcing, big(-ish) data and activist social analytics. The software ethnography gives us a rather rare opportunity to go deeply into the processes and logics by which this kind of data is produced, structured, represented, circulated, understood and used by the full range of stakeholders (whether women in a low-income neighbourhoods or municipal authorities being lobbied for infrastructural improvements). Above all, the informational logics built into the SafetiPin system can be brought into dialogue with the understandings of women in our fieldsites: what different understandings of safety, risk, urban practice can be identified and how can they be brought into better alignment?  What lessons can be learned for democratized ‘smartness’?

So this is a great opportunity to work on the connections between new technologies, women’s lives and urban systems. And because we are partnering with Safetipin, we can do this at a deep material level – looking at the assemblages of the software/system as well as at what women do with it. And above all, we can bring together the way Safetipin and its women users understand urban risk and safety.

Put more pragmatically, by combining software ethnography and end-user research we can understand and build on the fit between the software and its users.

I like the term ‘software ethnography’ (a word phrase that seems to be used, if at all, in productively divergent ways) because it applies equally to SafetiPin as an organization and as a software object: we are entering into the world of the people and systems that produce SafetiPin (as a campaigning organization, as a data generating system, as a software app and service); and we are entering into the life of the objects that it produces – how do its ‘safety scores’ circulate through diverse social worlds, interacting with different governance systems, human understandings, infrastructural conditions and so on.

In the latter respect there is another crucial element: we are also working with the Delhi-based computer scientist Rakhi Tripathi, who has the expertise to view these issues at the level of code and algorithm, at the level of the properly ‘technological’ object.

Software ethnography of disconnected infrastructures

For the project team, the issues are pretty classically ethnographic.

  1. Safetipin is a unique system that both generates and aggregates data flows from numerous sources, including sourced from its own users – it is therefore itself an infrastructure as well as connected to other infrastructures.
  2. It then involves coding and summarising and modelling that data for particular purposes – to assess and communicate risk, to provide a knowledge base for women moving in urban space.
  3. Safetipin then involves representing and communicating the results of this modelling to women users and municipal actors through the app.

So a software ethnography has to look at quite a range of stuff, from people’s actions and relationships to data structures and algorithms. And then tries to look at these dimensions over time, as a developing process.

In practice: this breaks down into a few fairly clear research questions:

  1. How are risk, safety and infrastructure conceptually connected in Safetipin’s work? E.g., in the idea of social audits and in background theorizations of gender and safety, what conceptual links are made in the software?
  2. How is data organizationally generated, processed, coded and otherwise structured and process – what kinds of connections and concepts of risk, gender and urbanity are performed through their work?
  3. How are connections and disconnections represented through SafetiPin? At the level of algorithms but also of interfaces, what social connections and models are produced?
  4. And can we bring together the software ethnography and the community level work to see how the versions of risk, gender and infrastructure in the software, and in women’s experience, can be more closely aligned.

This slideshow requires JavaScript.

(c) The Comparative Analysis Report on Safety Audits Done in Delhi, Bogota and Nairobi: Using Data to Build Safer Cities, Safetipin

Which brings me to the final but absolutely central point: this is a very close collaboration with Safetipin that will no doubt take many different forms and dimensions. I suspect we’ll end up with many different names for these different aspects – we’ll be doing a study of Safetipin, a study with them, there may be co-design at any of the different levels I’ve been mapping out. We’ll see. This is something Jo and I have been exploring in various fields – development, creative industries, and now – in our Configuring Light research group (please see her first blog post on the Configuring Light programme here) – lighting design and public infrastructure. The main thing is responsive, collaborative, organic exploration – we’re seriously grateful for this generous access, partnership and collaboration.


Slater, D.R. (2013) New Media, Development and Globalization: Making Connections in the Global South. Polity: Cambridge.

By Dr Don Slater, Project Co-Investigator & Associate Professor (Reader) in Sociology, Department of Sociology, London School of Economics and Political Science 

One thought on “Software Ethnography: Notes from the field

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.