Back in October last year, Deep Baltic featured an article on the Baltigram project: an interesting and original idea, as well as one that seems particularly topical at the moment – creating maps based on social media use, in this case, photos posted to Instagram in the three Baltic capital cities, as a way to visualise how people represent themselves and view their environment. The project was initiated by Spin Unit, a collective founded in Tallinn which uses what they describe as a metamorphological approach to urban planning and analysis – attempting to take full account of people’s activities in urban space and making use of the latest technology to produce unconventional means of picturing cities. In their own words: “we aim to hack the invisible city, leverage the secret strengths of urban spaces, and explore how the many characters, all the different layers of the urban fabric come together.” Deep Baltic’s Will Mawhood recently had a chat with Damiano Cerrone, who is one of the main drivers behind Spin Unit and is currently based at the University of Tampere in Finland, about their philosophy, their previous projects and where they see technology taking the subject in the future.

Damiano Cerrone

Could you tell me a bit about the background that led to you setting up Spin Unit? What the impetus for doing this? You’ve got an academic background…
At the very beginning, the main idea, or dream, was to have different practices from different faculties of this academy of arts [Tallinn] coming together to create one research unit that combines arts and science, in order to look at urban phenomena. So this was the reason, but practically what happened is that we had all finished our master’s degrees, and we were thinking – “what’s next?”. So in one way that was the goal of Spin Unit, and in another sense, the practical goal of Spin Unit was to start to build up a new research-based practice, since we were all finishing our master’s. And we didn’t find other companies or institutions that were doing that in Estonia.
We first proposed this project to EKA [the Estonian Academy of Arts] itself, looking both for moral and financial support to start up an interdisciplinary research unit, then we started looking into consultancies but the artistic approach was not really accepted. So we thought we would try by ourselves, using our savings to self-finance the first experiments and build our first sensor machine to collect data, sticking to the name of UNIT because we are not activists, but we are also not a consultancy. We wanted to study the city and reveal invisible stories and to tell good ones we needed to develop a research practice capable to work both in the fields of data analysis and anthropology. We believe that good stories come from the representation of quantitative and qualitative data together. Now we are no longer a not-for-profit but nothing has changed, our motto is still the same: “combining art and science to reveal the metamorphology of the city”. What’s changed is that we now have a a fully inter-disciplinary team of academics and practitioners working together in a virtual office across timezones.
I noticed that you have a number of people involved in Spin Unit who have an artistic background – you have, for example, a philosopher on the team. What does this give you that people who are purely looking at it in a scientific way don’t?
I think it’s about two things. One thing is about interpretation – the mindset of an artist is way more open to discovering different issues that remain hidden behind the Excel table of data, that I wouldn’t or other scientists wouldn’t. They see things in a more critical way; they wouldn’t just rely on the data but would go beyond this. The second thing about having critical thinking applied to data visualisation is that we also get an output, which is an interpretation of the data, not just its map. So one thing that we really work with is not just presenting information as a set of points on the map, but to try to present maps that can be interpreted and understood, even by those without a background in science. We try to show maps that don’t have a key, because we want to make sure those maps are understandable even without a key; so it’s data visualisation in art, basically.
Could you give me a concrete example where one of the artists involved has seen something that you haven’t?
Absolutely, yes. They still are contributing in that sense. One thing that we do is to analyse social media data and representing it on digital cartographies. But when you observe the actual content, you need a clearer understanding of that. A scientist takes this data and visualises it on the map, looking for patterns in the data. But what the artist does in this case is not looking at the patterns in the data, but looking for stories. Actually, I would say that this process is taking us away from being urbanists and more towards being digital anthropologists, or at least that’s where our interests are focusing right now. So it’s actually trying to retrieve these stories behind the pictures that people are publicly sharing. As we speak, we have a data scientist in Estonia, a physicist and an art historian in Toronto, two anthropologists and eighteen researchers working on data collected from over 100 cities. Searching for stories, revealing the invisible images of the city, that at the end I visualise in the form of maps, photo collages or dedicated webpages.
We’re obviously particularly interested in the Baltigram, because it unites the three Baltic capital cities. How did this come about as a project?
We were curious to understand to which degree the urban environment changes the perspective of one self. So we started to study selfies taken in the three Baltic capitals to see if people represent themselves differently in selfies depending on their location.
Images from the Baltigram project

As you mentioned in the text you wrote for us, the Baltic countries are quite different; even though they’re often seen as a unit from outside, when you’re here they feel very different as countries. But the data you got was broadly pretty similar for each capital city. There were certain interesting variations that I noticed, which we can talk about in a moment, but basically they were quite similar. But do you think that this means that they are similar at some level, or does it mean something different?

I think there are so many different answers and interpretations, but what we have seen so far is that the Baltics share history, culture and consequently a certain type of architecture – so, for instance, the old towns, the Soviet blocks and these post-1991 single family house estates.
Although the architecture itself would be quite different in the Old Towns – between Tallinn and Vilnius, for example.
It is true that it is different, but in many ways they were built at the same period or have shared characteristics. All three Baltic capitals were part of the Hanseatic League so they have more or less the same influences, although reconstructions and changes have occurred in modern times. Old towns are also places for tourism and visitors to these spaces commonly share selfies to strengthen their presence in iconic places by including most known architectural features in the background of their photographs. The Old Towns are touristy places and places for social interactions, so many people go to there to meet with their friends or to have drinks. Of course, this is changing right now with the rise of new informal places like Telliskivi, but it still has these iconic elements and gathering points. If someone is in Tallinn, it’s a must to take a selfie in the main square (Raekoja plats) just like nobody misses Senate Square in Helsinki.
But the point is that this kind of architecture becomes almost like a point of pride, or it becomes a means to locate yourself, to show someone that you are in the Old Town and gain an expected amount of likes. Receiving a like on social media triggers a dopamine high, just like using cocaine or methamphetamine, and if someone knows that taking a certain picture with symbolic elements in the background will trigger more likes, then they will be more encouraged to take easily recognisable pictures that are the most likely to receive likes. Therefore there are many readings and causes for this phenomenon, even the factor of replicating pictures taken by social-media VIPs or just attraction to certain architectural features that are common throughout the three cities, because they all have some kind of iconic architecture in the background. All in all, we always have to consider that in social media local knowledge and global trends are causes and consequences of what pictures we take and why.
It occurred to me though, I’m sure this you’ve considered this, but it’s quite likely that the people in the Old Town, in particular, and the Soviet blocks in the suburbs, are different people. The people in the Old Town are much more likely to be visitors from elsewhere. What people notice, particularly in Tallinn, maybe not quite so much in Riga and Vilnius, is that locals don’t go to the Old Town that often – maybe in the evening. Does it necessarily tell us something about the people who live there, or does it tell us something about how people outside the city see it?
Exactly, it is a bit hard to separate the tourists from the locals. We could, but we follow an ethical code banning user profiling from our research. When we collect data we don’t store users’ name and info, although this is allowed and companies like Cambridge Analytica use it not just to study but to influence behaviour, as in the well-known cases of the Brexit and Trump campaigns.
What we can say is that the type of environment is what influences the way we interact with it. If an environment is iconic and beautiful, we always try to bring it inside the picture. But in the case of the Soviet blocks, one of the thousand reasons – of course, there are a thousand reasons – but the reason why most of the pictures are indoors, in very closed spaces, is perhaps because in these areas, architecture doesn’t play an important role in people’s life – or even more, it’s purposely left outside. The person become the most important element in the pictures. A huge number of pictures are taken in elevators there, because there are mirrors. The mirror kind of helps to take the picture, because the elevator gives this moment of intimacy. The mirror is a medium for oneself and the lifts in the Soviet blocks turns into a popular space for privacy, but the buildings are never there [in the photographs]. Local researchers and anthropologists can use our data to undertake more in-depth research regarding this phenomenon but our goal here was just to find this story, to show this invisible image of the city that later, with all the data we collected, can be analysed to draft further conclusions.
Did you see any exceptions where people did seem to be highlighting the Soviet blocks? Or were these almost invariably excluded?
No, they were basically excluded. The only few places we saw this architecture in the background was when teenagers were hanging out in the empty courtyards, or sometimes if there was some kind of urban landscape [in the picture]. So the Soviet blocks were highlighted in very few pictures, of course, mostly in pictures taken from bridges – in Tallinn, for example, you would get a picture from a bridge, so you would see the entire landscape of the Soviet estates. So the blocks seem to be more recognised and accepted for their urban landscapes rather than as architectural elements or courtyard activities.
I wanted to ask as well about the photos taken in the suburbs, because in the Baltic countries, in the capital cities at least, if you have your own house, this indicates you’re probably doing better than the same thing would in most Western European cities. It seemed to me that the people in these pictures were showing off certain kind of prestige possessions – I don’t know if that’s a trend you noticed as well.
Yes, absolutely, that’s exactly the trend we saw. It’s about highlighting private property, showing all the activities going in private backyards, showing the greenery. In fact, if you look at our collages on the Baltigram website, you will notice that the colour green is dominant in all pictures. One thing that was excluded, just because we wouldn’t want to show them in public, was toddlers and kids. A massive percentage of pictures had toddlers and kids them; in the backyard, still in this context of family meetings at the weekend, and we decided not to show them according to our ethical protocol. But yes, again, it’s about private property and owning that little space of nature in your backyard, and being proud of it. Something similar to the American dream but shared throughout the Baltic states.
Do you think in a way you could reduce the difference between the Old Town pictures and the other two types to “where I am” versus “who I am”?
Yes, I definitely think that in the Old Town it is about the location, about “where I am”, using the architecture to put yourself in context. In the Soviet blocks, it’s about myself in my own private space, in my own room, in the elevator. And in the suburbs, it is about the space I am in because I own it, showing that “I made it”. The space doesn’t represent me, but it’s me showing you what I own, showing you that I have this little garden. So those are the three main differences in that sense if you look it from the perspective of the relation between people and space. There are so many other readings we can come up with and that’s what fascinates me the most about publicly shared pictures.
Do you think that if the Soviet blocks were more – well, it’s subjective, isn’t it? – but more conventionally well-designed, people would be more willing to take pictures with them as a background, or do you think it’s the nature of being in this kind of area of the city – a comparatively poor area – it’s irrelevant?
That’s the thing, there could be different aspects and all could be true. One aspect could be the image of the city. If the architecture of these places was perceived as iconic, for historical or architecture quality reasons, the general understanding of this space would be completely different. It is also about activities. When one space, regardless of its architectural quality, is inhabited and used by people, it becomes automatically recognised and it draws more people feeling proud sharing their locations in those places. I’m talking about the early Telliskivi, the informal places along the bay of Tallinn, but also people doing parkour in decaying areas or drinking pop on spot. If the architecture itself didn’t carry the connotations of a hard past, of being a difficult place to live, maybe it would be more present in the pictures regardless of the look or conditions of the architectural structures themselves. What I’m saying is that the perceived images of the blocks play a bigger role that their actual design and physical condition. So that’s the difference – we cannot say for sure if people don’t show themselves outdoors or in front of a building or in the courtyards, because of the buildings themselves or because they simply don’t meet outside due to cultural norms. That could also be one reason: we might never see the architecture, not because people have decided not to include them in the picture, but because people do not hang out in these neighbourhoods, but go to the city centre to meet with their mates.
Another reason we found in Russian cities is that teenagers avoid courtyards because when they are there they can be watched by their parents and by babushkas watching from the windows. The architecture of the mikrorayons imposes a very strong level of natural surveillance, so much that may keep people away from hanging out in the courtyards for the lack of privacy – which instead one has in the single family houses backyards.
Let’s think about the Old Town shots. Spin Unit is mostly involved in what you’ve described as “interactive urban-based design”. How could photos of the Old Town, which is essentially a tourist site, help an urban planner or someone involved in urban design in some way? What practical insights could you take from that?
Recommendation for improvements are site-specific and tied to the goals of one project. When using big data one can get interesting answers only when one asks precise questions beforehand. It’s hard to generalise. If we used images to study what the most attractive places in one city are and shared that information with municipalities we would trigger very dangerous planning practices. Cities would start copying each other or even more municipalities might make the wrong investments, in places that are already popular and well-recognised. So what I’m saying is if when in consultation with one developer, I tell him “OK, Tallinn people prefer those kinds of shops or those kinds of urban environments”, redesigning the same environment in another part of the city or in another city like Tartu would be a massive mistake. The reason a place becomes attractive is due to a complex combination of design, culture and practices; I’d be completely against the idea of trying to replicate that. But this is what is happening at certain levels – so unfortunately that’s the reason why you see urban developments in different parts of the world looking alike: it’s because they’re trying to replicate the success of popular places in other cities.
What you can do, and what we do – I’ll give you one very concrete example: right now, we’re doing a project in Russia, and what we can clearly see from the pictures is that there are a large number of teenagers spending their leisure and social time in former industrial areas – so, in dangerous places. So the idea is that then the city has to decide how to change this interaction. Are we going to improve the interaction – meaning that: OK, teenagers like this area, so we are going to make it safe for them – or are we going to close this interaction? Do we build a wall around this industrial block, so that the interaction is erased and cut out or do we recognise the people’s attachment to one place and help make it better? These are the questions we are working on at the very practical base and decision will always have to be taken according to local conditions. Never use this data to make generalisations; that would be the biggest mistake.
Another example is one in Tallinn – the other quick investigation we did. We saw many people discussing about the construction of a new road in Tallinn, cutting through a very popular beach in Tallinn, reshaping the shoreline. Politicians were saying that this area of Tallinn was not that popular and mostly characterised by anti-social behaviour but we knew it was not like that, so we started to search into social media data to reveal another face of the invisible image of the city. There were thousands of pictures proving the place was extremely active and had a strong social-cultural importance for local communities.
This is the area around the memorial?
The Russalka Memorial, yes. According to the politicians and common opinion we were scanning in the news, there were not many social interactions in this part of the city, unless they were bad ones. So, based on an idea from Kristi Grišakov, we built a team overnight and started to work on this data, which showed a different story from the one told by politicians – we showed that, like any part of the city, there is anti-social behaviour, but there is also something quite beautiful. There is a very strong connection between people and this area, and it’s used throughout the whole year. So that’s another thing – you can see from looking at the pictures that people do leisure, they do sport, there is attachment. Many cities are struggling to produce attachment to new places in the city, they invest millions in place making and making new development more recognisable. Tallinn had it for zero cost; it had the whole area already socially developed and symbolically recognised for zero cost, it had a sandy and rocky beach in the same place for zero cost but they are spending money to get rid of it altogether.
And that was using social media was it, again?
In that case we used Instagram but we have data from many other sources too.

Instagram pictures from around the Russalka memorial [Image: Spin Unit]

One possible criticism I could think of is: how are these examples different from surveillance? The example you gave of teenagers – OK, they’re putting it into the public sphere, but they’re not necessarily doing that so that people can watch them and see where they are.

Yes, that’s correct. I mean what we do, the way we operate is that we do not track single users. It is possible – it is possible to track single users, but we don’t collect that string of data when we download pictures, so we don’t know who took them. Consequently we also can’t track the movement of single users. The other thing is that we take a random sample of pictures. This means that, for example, in the city of Tallinn there were maybe 15,000 Instagram pictures from around the Russalka memorial, but we take a random 5,000 pictures sample for our study. So in this way we work within our ethical code and avoid getting even close to gather personal information.
So you’re not following individual people?
No, we’re not. Because we’re not interested in that. And the researchers who study the images are using a custom software we developed which shows the images but not the location or time. They only see a picture and they classify it according to our method.
By using social media to base policy or recommendations for planning on, are you not excluding large numbers of people? I’m assuming you have some kind of filter, but I’m not thinking just of older people, but also certain kinds of personalities? Certain kinds of people, regardless of their age or anything else, they’re just more likely to post on something like Instagram.
Yes, that’s right. So actually this is in many ways an advantage, because we have a clear image of what is our sample, of what is our population. Depending on location, our sample is mostly people from 14 to 35 years old, but in Russia only half of the pictures with people were with teenagers. Of course the sample is changing year after year as more people of different ages start to use smartphones, but basically we have a clear image of what is our sample, what is our population, which allows us to combine our research with pre-existing surveys that generally reach middle-aged and elderly data – because they are the ones with more more time to spend answering questionnaires and usually the older people are the more stable and traceable their activities through the services provided by one city. On the other hand, teenagers and millennials are the one shaping the new society, with hard-to-trace habits, non-routine and fast-changing social behaviours with changes that occur too fast for a questionnaire to detect them.

So it seems that – and I’m actually taking this from something that Daniel Giovanni, your colleague from Spin Unit, who gave a talk which I watched. He was talking about the importance of particular lines of movement that don’t follow what has been planned – such as where on an area of ground people have cut across rather than following the set path, also perhaps what sometimes happens at traffic intersections. And he said “if this happens, this is where the urban planner failed you”. Do you think the easiest city is the best city, in this kind of sense, or are there examples where the planners do know best?

First of all, what we mostly study is where people take different paths from the ones that are designed beforehand, or where people organise space differently from how it was designed. What we think is that the main problem is that planners most of the time know what to do, because they are trained for that. But the problem is that the planning rules very often don’t allow them to make such big changes. For instance, it might be that you may have a classical design of a square – “classical” meaning that it was maybe designed 50 or 60 years ago – and planners know it must be changed because people do not move in that way or do not walk in that way anymore. But of course very often it’s nearly impossible for planners or designers to change the squares that quickly. Even if you think about squares like the one in front of Viru Keskus in Tallinn [a shopping centre in the middle of the city] – it’s taken years to change the layout.
We are interested in margins and alternative behaviours. What we see as the main problem in modern cities is the authority of cities that rely only on long-term land-use-based planning. In other words, cities that are designed just based on the ownership and function of one place while society changes at a faster pace than their plans. But the problem is that the planning rules very often don’t allow for fast changes or instruments that allow temporary uses.
I think that urban planning systems – meaning laws, policies and regulations – are outdated. Think of Italy, where the national planning law is still based on principles dating back to the Fascist era and a unified urban standards regulation made in the ’70s is still in force today. Planners just follow the rules, the rules are failing to follow changing needs and as a consequence, people have to self-organise to accommodate new practices. New practices can arise in the blink of an eye but planning instruments take years to react.
I think that the more a city is open to accommodating new practices, cultures and ways of life, the more successful and attractive it will be.
It occurred to me though that if people take a shortcut, this is not always the fault of the planner. I mean, maybe pedestrians are taking a route that is dangerous, or just is ugly in some way – maybe organising the space in that way would be aesthetically displeasing.
Yes, of course, I completely agree with you. That’s just a small example for the general public, but usually when we actually work with city planners what we do is that we are mostly trying to work with land use. Planning fails when people are trying to change the use of a space more quickly than the planners are. You even see this in areas like Telliskivi – people were trying to change the space much more quickly than the planners would have, because in the planning scheme, that was an office area. But of course, people have accelerated the process from an industrial area to a cultural space. So what we are looking at mostly in this case, is that people really need a different function to be established in one area, and in many cases, people actually occupy or squat certain areas, to change the function or to accelerate the process of change. But then the planners fail them, because they are not taking these shortcuts.
Where do you see this going in the future? Do you see social media continuing to develop in a way that enables you to track and use it? Do you see this becoming more central to urban planning?
I think it’s becoming more central if you want to get a sense of place on the large-scale. So if you want to understand what the general patterns are, what the general activities in the city are, and if you want to understand what the stories of those cities are. But about the future of using social media data… I’m not sure. The fact that we are doing this now, in the present, it means it is not the future, so we need to put the same amount of time in refining our technologies to develop new ones. We’ll keep investing in our resources, developing our methodology and perfecting our tools to the same degree as studying new fields of enquiry in order to find the next big thing that is interesting for us.
So far, social media data is great for understanding general patterns and the story of the city, but we still need to engage in the physical space, and really get a sense of the community, especially those who are not on social media. I can’t say how central it will be but all projects I’m engaged with are about using social media data to draft new development plans.
So that could be fed into some kind of algorithm and you could have insights as it happens?
Yes, and that’s actually what we’re working on right now. And the reason why cities want that as a service is because they actually are aware that participatory planning is happening, and they believe that people are really participating in planning in the traces that they leave. So every time they take a shortcut, every time they do something different, they are changing the shape of the city – that’s kind of the idea. So the city wants to see these phenomena as they occur, not because they want to monitor people, or because they want to track people or stalk them, but because they want to make it a better city for them. Of course, it is extremely dangerous, because it really depends who is using this information and how they are using this information. Because again, if you want, you can track single users; if you want, you can do that.
All images credit – Spin Unit
You can find out more about Spin Unit’s work throughout the Baltic region and beyond at their website. For more information about the Baltigram project, take a look at Deep Baltic’s feature on the subject.
© Deep Baltic 2017. All rights reserved.
Like what Deep Baltic does? Please consider making a monthly donation – help support our writers and in-depth coverage of Estonia, Latvia and Lithuania. Find out more at our Patreon page.