What do architecture, investigation, maps and storytelling have in common? Follow the trail in this episode where Alison Killing takes us on a fascinating journey starting from her training and professional career in architecture and constructions to her research in migration and human rights, culminating with an internationally awarded journalistic collaboration.
That wasn't just a problem with the map loading. And it wasn't just that there was information missing from the map and they put a blank tile instead... there was something much stranger going on. I mean, I find it quite enjoyable to push and poke at these different things and try and develop these different investigative techniques and ways that we can maybe start to look at these issues, which are by design very, very difficult first to investigate.
About the speaker
Alison Killing is an architect and urban designer who increasingly works on investigations and digital storytelling. Her work centers around issues which have a strong spatial dimension, such as migration, street level surveillance in public spaces, or the construction of networks of prison camps. Alison's most recent work has focused on documenting the scope and impacts of China's system of mass detention for Muslim minorities in its Xinjiang region, and her team's series on the subject, which used geospatial analysis, architectural modelling and dozens of interviews, won the 2021 Pulitzer Prize for International Reporting among other distinctions.
My name is Alison Killing. I'm trained as an architect. In fact, I am an architect, licensed in two countries. And after I graduated in that I worked within the construction industry for a number of years, for structural engineers, for architects, for urban planners. And then I left that in 2010 to start working for myself.
I had the ambition to build when I started working for myself, but the economy was very weak at that point. So there wasn't a lot of building going on, regardless of who you were and how experienced you were. And so I wasn't building and I started to do other things instead, and I worked on research projects and community design and build projects. I did exhibitions. And then I started working on an online documentary about migration to Europe. And that was really the point at which my work shifted properly towards journalism. Since then I've been working - yeah, doing sort of investigative projects using, often using my architecture skills. And we've been looking at the network of camps in Xinjiang and I've been working with a team at Buzzfeed over the past two, three years to do that. The migration work had been focused a lot on storytelling and how to tell this story effectively, whereas with the work that we did on Xinjiang, it was really... the challenging thing was to find out what it was that was going on in Xinjiang, and so investigation was much more to the forefront in that.
2016 was a pretty tumultuous year, for people in the UK and in the US I guess, what with the Brexit referendum and then Trump being elected. And it just started to feel like, that there were a lot of quite urgent social issues that we needed to deal with and they needed an urgent response. And as much as I think that architecture and urban planning are really important - you know, I still think that they're really important - but they're also incredibly slow. I mean, I used to work on like fifty-year strategic master plans. It's the sort of thing that you might not even like, live to see them completed. And so I was drawn to work that unfolded on a slightly - well on a much shorter - timescale in fact, and that had the possibility of impact with that.
And so I was kind of looking for opportunities to get involved with that. And they're quite difficult to come by. Like, I don't have the sort of skill set that is commonly represented in newsrooms. This isn't a common or mainstream skill to find in newsrooms, that I could just apply for a job doing the thing that I wanted to do. And yet I could see that I had skills that could be useful. And so I was looking for opportunities that went with that, and I guess the first thing that happened was that I saw the advert for the first Exposing the Invisible workshop - and I saw it and I was like, oh man, that looks really cool. But you know, like I just, I don't think I'm qualified. They're looking for journalists and investigators and people like that. So yeah, I got accepted to that workshop.
I guess it was just like the first time that I was amongst a group of experienced journalists and investigators and being taken seriously amongst that group of people. Also just to be able to sort of talk with people and show them what it was that I did. And to help me understand like what it was that people - what they understood my skills to be and what they sort of understood that I was able to do, which was actually much less than I had initially thought. I think that a lot of the work that has been done around investigation and with architecture had been, up to that point, really quite complex projects around modelling and animations and sort of reconstructions. These really very precise, very forensic reconstructions of events. And I was like, okay, that's what I need to explain to people: this is how you do it, this is what the opportunity is with this, with that sort of work. But in fact, what I found was that people often needed things, which I had considered initially much simpler. You know, accessing historical satellite imagery, and how you can use that and how you can access it. From understanding what other investigators needed, from understanding where they were, what they understood about what I did and what it was possible for someone with my skill set to offer, it opened the door to... yeah, that actually opened the door to future collaborations. Megha and I met at that workshop.
It was believed that there were about 1200 camps in existence. Around seventy of them had been found. Megha had been the first journalist to visit one of them. And shortly after doing that work, she'd lost her visa for China. So she no longer had access to China, and yet was still very keen to continue working on this. Access to Xinjiang was difficult for any journalist who was trying to work in China. And kind of for two reasons. I mean, the first is the fact that Xinjiang is just so big. And so it's a huge amount of ground to cover, to try and find all of these camps. But the more important issue is actually the Chinese government's control of information or attempts to control information and the harassment of journalists who want to work in Xinjiang and also like harassment of sources. People are very reluctant to speak to journalists because they fear, quite rightly, that they could be sent to camps if they do. You know, they were reporting that for people who are trying to work in Xinjiang, the Chinese police would be staging car crashes to stop journalists travelling down certain roads to try and visit camps. So it was incredibly difficult to work there. These challenges were something that satellite imagery was able to overcome because it's something that is very, very difficult for the Chinese government to control. It can also deal with the issues of scale. And so we realized that, yeah we had this complimentary skill set and we could potentially work together quite successfully to try and find this network of camps.
I definitely had some experience with satellite imagery and I had skills which allowed me to... so, I mean, one thing is that, as an architect, what you tend to be - well, in fact, you have to be very good at thinking in three dimensions - and you have to be very good at moving between two-dimensional representations of objects and a three-dimensional version of that, whether that's a representation or the object itself. And so having constantly over the course of my studies, and over the course of my career, been working with, you know, plans and satellite images and elevations and drawing sections through buildings, it was a very very short step from there to be looking at satellite imagery and interpreting what we were seeing on the ground, to being able to look at photographs taken on the ground and to be able to match them to satellite imagery, what we were seeing there and identify the different elements. That was very straightforward.
I had two screens set up and on one of them I had Google Earth. And on the other one, I had a tool which Christo Buschek, the developer who worked with us, had built. And that had these point locations of all of the places where they were mask tiles superimposed over Google maps. And I was just going through all of these points going like, 'no, no, no, no, no, no, no'. Like, maybe that's the thing. And I'd zoom in a bit on Google Earth and be like, 'no, not a thing. No, no, no, no'. And then occasionally a yes. Or occasionally, rather a maybe. And that, those markers together built up a database, which that database then formed the real backbone of the project.
We mapped out all of those mask title locations. And it was a bit of a shock to find that there were five million. I had thought that it was maybe going to be somewhere between 500,000 and a million, which is still actually a lot, but five million was, you know... it's ten times bigger than what I'd expected. You sort of realize that actually you've got ten times more locations than you had originally expected and you sort of do this sharp intake of breath of like, 'oh God, I hope we're going to be able to like, still work through all of this, how are we going to get it done?' And then you take a deep breath and you're like, 'no, of course we're going to manage to get it done. This is the job. This is just what we do'.
And then you start working through it and you find the ways that allow you to narrow that down, which in this case was to say, we're going to focus on the areas close to towns and the areas along major infrastructure. In that way we managed to get it down to like fifty thousand mask tile locations that we needed to look at, which is still a lot. But I went systematically through those locations. I was managing to do like ten thousand in a week, which is quite a lot. And that was like quite an intense week. But it's also, it's manageable, like five weeks of work is... we're going to get there, we can see the end of that when we start.
When we began, there were about seventy camps which had already been found. And so we knew quite a lot about key characteristics of those camps. One thing that I found though very quickly - and I remember calling up Megha to tell her about this, cause it was such... it was like such a shock to me, and it was like, 'oh my God, I have to tell you this.' We had expected to find about 1200 of these camps, but after even just a short amount of looking and trying to identify these locations, I realised that that almost certainly wasn't true. In the beginning, the program had really been a series of government-owned buildings - schools, hospitals, maybe empty apartment buildings around Xinjiang, which had been taken over and then in a matter of weeks or even days had been converted into prison camps. And what you saw with those camps was... the giveaways with them was barbed wire fencing in the courtyards. You would see these passageways running between the buildings lined with barbed wire. You would see these blue-roofed industrial sheds springing up on the playing fields or squished in between the buildings. And you would see these impromptu car parks, which appeared just outside the site. And then with the later ones, they were purpose-built, they were much higher security, much more intimidating facilities. And they started to look quite different and they started to look much more like prisons which you can see elsewhere in China. Those facilities would have these really thick perimeter walls - really thick would be like a metre and a half to two metres wide walls going around these places. On either side of that wall, there would be one or maybe two layers of barbed wire. There would be guard towers at the corners and at regular points along the wall. Because those places looked very similar to compounds elsewhere in China, also because we had in some cases tender documents or media reports, we were able to corroborate, with quite a high degree of certainty, that in fact a large number of these places were camps. And then for the rest, we could basically look at the similarities between camp A, for which we had a load of corroborating evidence - we maybe had an eyewitness statement, we might have a tender document, perhaps a journalist had gone and visited - and then we would see exactly the same features in compound B. And we could say 'okay, compound A we know as a camp, compound B has the same features: we can be reasonably certain or we can say that that is also a camp. And we can say that with a very high degree of certainty. And then going to Megha and being like, 'oh my God, I found this', or 'I found that'. She was then working a lot with ex-detainees, doing these really detailed interviews that helped us to corroborate a number of locations. But it was also about understanding what the human story of this, of this issue, was. What was going on more widely in Xinjiang? How did it feel for the people who were involved? What was the human cost of what was happening, of this massive detention programme and programme of oppression? And Megha did this amazing job of packaging this as like, 'how can we tell the story effectively?'
We were able to make a 3D model of one of the camps. Sort of give this detainees' eye view of what life was like inside the camps. Working back and forth between making this model, which I did based on satellite imagery and then a lot of quite basic architectural knowledge, I put together a draft of this model. And then from there, we were able to start asking questions of the ex-detainees who had stayed in that place about, you know, trying to find out how big the cells are. Because we were able to count the windows along the facade, we knew what the basic structure and what the structural grid was of the building and what a likely division inside was, but we didn't know how big the actual cells were. And so we were asking questions like, 'how many windows were there in the cell that you were kept in?' And it was like, 'oh, there was one'. And very quickly then we were able to narrow down how big that room must have been, and then how many people were kept in there. And there was actually a huge amount of overcrowding in that place that we were able to point to based on that reconstruction. Similarly, there were structures in the courtyard of this prison - of this camp rather - these like wire pens. And we were able to go to the detainees and say, 'look, there's these wire pens in the courtyard, what's going on there? What happened there?' And to have quite a good conversation using that model and using the questions that had come up in the course of my modelling to sort of say, oh, actually there's a series of different exercise yards for different categories of prisoner. But in fact, people were very rarely allowed to go into them. They went out maybe once in six weeks. And so this model served as a really good basis for a discussion that Megha was able to have with the detainees and to find out things like, they actually hardly ever left their cells. They were incredibly confined and incredibly controlled within those places, only ever moving about like 10-15 metres from where they slept. Even on days when they went to the classrooms, never feeling fresh air, this incredibly claustrophobic, incredibly controlled experience. And that was something that we were able to get to using this model and the techniques that I had and then Megha's skills in and experience in interviewing.
Migration Trail is a mapped data visualization. And it follows two fictional migrants who are travelling to Europe in real time, over ten days. And then we had worked with writers from - one writer from Nigeria, one writer from Lebanon - to write the voices of those characters as a social media feed, which the audience could get either on the website itself or on Facebook Messenger. And then we also made a podcast series that looked at the wider issues around migration, which was non-fiction. I've actually been interested in migration and refugee issues for a long time. I've got a master's degree in humanitarianism and development practice, as well as in architecture. So I'd been seeing, since about 2013 and into 2014, that there was quite a lot happening in terms of migration to Europe, in terms of irregular migration to Europe, or a big increase in the number of people who were starting to make this journey through the central Mediterranean, from Libya to Italy. And yet it felt like this issue wasn't getting the sort of media coverage that it actually deserved. And I remember reading something in the Guardian that said that even when they were writing the articles, reporting what was happening, those articles often had the lowest traffic on the webpage. So it wasn't being read. And so what I was interested in was trying to find a new way of telling that story and doing it through maps and data and telling it in real time seemed like it might be a good way to do that. The idea that telling it in real time would help to make it more immediate for an audience. The other thing that I should say that we tried to do with that was to move beyond just individual stories of the people making these journeys, and start to talk about the political context. And to start to talk about European border policy and asylum processes, which forced people into these journeys in the first place. It took two and a half years from beginning to end, although I wasn't working on that full time. Over that period, you don't really have a news focus, per se, because you just, over that timescale, it doesn't work to have that sort of focus. The advantage that it brings is that you can actually step back and provide this big historical context and this big policy context to what's happening. And that was what we felt was the strength of what we were able to do, stuff that it's actually - it's very, very difficult to provide in news reports - I mean, for issues of space, if nothing else. I would say that probably the storytelling came to the fore much more in Migration Trail. It really felt like a design project to me, like, okay, we've got this story and the story did need a huge amount of research, a lot of travelling, a lot of interviews, that I did. I sort of like travelled all of the major routes into Europe, right back up into Northern Europe. I spent two weeks on one of the rescue ships in the Southern Mediterranean. So there was a huge amount of research that went into it and it was really about, like, what's the most effective way to tell this story and what are the different aspects of the story that we want to tell, and how can we... what are the different formats that we're using? What are they good for? What can we do well with those? And then how do we work, how do we get them to work together to tell this really effective story? You know, the map was really good at showing the scale of what was happening. It was also really good at linking together what I think for many people were like this really disparate set of stories that were spread across Europe: about fires in the camps at Calais, about large numbers of people landing on the beaches in Lesbos, of a border wall springing up through the Balkans, and sort of showing how they could be tied together into one much bigger story. The social media storytelling that we did was really good at showing the personal side of what it meant for the people who were undertaking these journeys themselves. You know, giving people this very intimate picture of what it was like to do that journey and to be in touch with somebody who was taking that journey, because you were getting these messages in this very personal intimate space on your phone, and in a messaging app where it really did look like the messages were coming just to you. And then the podcast was also like a really nice format to work with because there's a number of things that you can do quite successfully. Audio is a very good medium for discussion of quite complex issues. So that allowed us to really get into the policy, the history of what was happening with migration to Europe. It's a medium that works very well for carrying emotion. So when somebody was telling us about, you know, the journey that they'd just gone on, it could really successfully convey what had it been like for them to do that and the emotion that had been involved in that. And it's also a very good medium for recounting historical events. So somebody could be telling us about something that had happened two months ago, when they were leaving Syria, up to something that had happened yesterday when their boat sank and they had to be brought in by the Greek coastguard. And that could still work in that format and could work very successfully within that format.
I've been trained in design and then also in making things. Obviously there's a creativity there, but you also have to really understand techniques very well and what to do with them, that ability to develop concepts and then to come up with concepts and develop them rigorously. That helped a lot with the Migration Trail work that I was just describing, where you have a story and you're trying to work out how all of these different elements fit together successfully. That to me is really, is very much a design problem. I really enjoy that combination of creativity and sort of understanding of technique that needs to go with it while you're trying to understand how it is that people are working with a given technology, or what it is that a given technology does. Because of the Migration Trail work I had this really good understanding of how interactive maps work. And that allowed me to know that when I went to Baidu and I found these strange blank squares in the locations of these camps, that that wasn't just a problem with the map loading. And it wasn't just that there was information missing from the map and they put a blank tile in instead, but that there was something much stranger going on. I mean, I find it quite enjoyable to push and poke at these different things and try and develop these different investigative techniques and ways that we can maybe start to look at these issues, which are by design very, very difficult first to investigate - to look into what's happening in Xinjiang when the Chinese government has been working very, very hard to shut down access and shut down access to information. And yet, we've been able to find these ways around that.
I'm really enjoying what I'm doing now, but it's just really difficult to say what I'm going to be doing in ten years and where I'll be. You know, ten years ago, I was convinced that I was going to be an architect and building buildings.
I do now feel that I can more confidently call myself an investigator and refer to myself as a journalist. I feel like I've proved that I can do this now, and I have the right to say that about myself. I think I just don't know yet in what sense winning the Pulitzer has had an effect on our work. When the announcement was made in the middle of June, we were right in the middle of this story where we were trying to calculate the capacity of all the camps. The following week, we were still busy with that. The work just goes on. There was plenty of work first to do. I mean, it has been amazing to get that recognition. I think for me and Christo, there's definitely a sense that the different skills that we bring are valid and useful and have really very real contributions to make to the field of journalism and to investigation. But neither of those are traditional journalistic backgrounds. And so it was great to have those techniques and approaches sort of recognised and validated in that way. But it's lovely, obviously.
The Exposing the Invisible podcast series is produced by Tactical Tech.
Interview, production and sound design by Jo Barratt.
Tactical Tech's Exposing the Invisible team includes Laura Ranca, Wael Eskandar, Marek Tuszynski and Christy Lange.
Music by Wael Eskandar.
Additional music is Headway by Kai Engel used under a Creative Commons 4.0 Attribution licence and Machinery by Kai Engel both used under a Creative Commons 4.0 Non-commercial Attribution licence.
Illustration by Ann Kiernan
This podcast episode is part of a series of resources and publications produced by Exposing the Invisible during a one-year project (September 2020 - August 2021) supported by the European Commission (DG CONNECT)
This content reflects the author’s view and the Commission is not responsible for any use that may be made of the information it contains.