In this episode of the PolicyViz Podcast, I speak with Yanni Loukissas, an Associate Professor at Georgia Tech, about the importance of context in working with data. Yanni argues that data are not universal but local—shaped by their origins, environments, and the intentions of those who collect them. Our conversation dives into how this perspective challenges dominant narratives in data science, particularly the assumption that datasets are neutral or universally applicable. We also discuss how design and storytelling can play a role in exposing the situated nature of data and how educators and practitioners can better teach and communicate these ideas.
Resources
Check out Yanni’s website and read his book, All Data are Local
Guest Bio
anni Alexander Loukissas is Founding Executive Director of the Interdisciplinary Media Arts Center and Associate Professor of Digital Media in the School of Literature, Media, and Communication at Georgia Tech. His research is focused on helping creative people think critically about the social implications of information technologies.
His most recent book, All Data Are Local: Thinking Critically in a Data-Driven Society (MIT Press, 2019), is addressed to a growing audience of practitioners who want to work with unfamiliar sources both effectively and ethically. He is also the author of Co-Designers: Cultures of Computer Simulation in Architecture (Routledge, 2012) and co-editor of The DigitalSTS Handbook (Princeton, 2019).
Different Ways to Listen to the Show!
New Ways to Support the Show!
With more than 250 guests and 10 seasons of episodes, the PolicyViz Podcast is one of the longest-running data visualization podcasts around. You can support the show by downloading and listening, following the work of my guests, and sharing the show with your networks. If you’re interested in financially supporting the show, you can sign up for my Patreon platform, make a one-time payment via PayPal, or shop one of the show’s sponsors.
Transcript
00:02.11
Jon
Hey, Ani, good to see you. Welcome to the show.
00:04.74
Yanni
Thanks so much for having me, John. It’s great to be here.
00:07.51
Jon
ah Very excited to chat with you. We’ve got some good stuff on our plate. We’ve already talked for like 30 minutes before we recorded, which is always a good sign. um So kind of two main areas I want to talk to you about. One is your your book right here.
00:31.28
Jon
um And then I want to talk about some of your your of your current work. So um maybe just a ah quick start of, for folks ah who you are, where you are now, and then we could talk about like, we’ll just start with the book, like um what inspired you to write it and you know, what do you hope people get out of it?
00:49.57
Yanni
Yeah.
00:52.48
Yanni
So yeah, I’m an associate professor of digital media at Georgia Tech. I’m calling in from Atlanta, Georgia. I’ve been here for about 10 years. I teach within the ah College of Liberal Arts at Georgia Tech, Ivan Allen College. And then we have a very interdisciplinary school, the School of Literature, Media, and Communication. So it’s it’s It’s a bit of a ah eccentric grab bag of people from across the arts and humanities and design, um but um a very fun place to work.
01:29.56
Yanni
i I really started writing the book in earnest when I got here.
01:30.26
Jon
Mm hmm.
01:35.25
Yanni
I had been in Boston previously, and I had been working on database for quite some time. But upon arriving in Atlanta,
01:46.70
Yanni
You know, with this new job, I really wanted to craft a ah larger scale project for myself.
01:55.08
Jon
Yeah.
01:55.27
Yanni
And this was, let’s see, I arrived here in 2014 and this was like the era of big data. ah You know, everybody was talking about big data and and you don’t hear that phrase so much these days.
02:02.58
Jon
yeah Yeah. No, that’s right.
02:07.95
Yanni
Of course, it’s in everything we do with generative AI and so forth. But ah it was novel and it was, you know, we we we were seeing these streams of data at a scale that was unprecedented and people didn’t know how to think about it. um And there was this sense that, you know, big data, we’re going to change everything, you know, education, research, government, business. ah And ah I was interested in
02:40.68
Yanni
finding some creative and critical ways to think about and work with big data. And the more I got into looking at big data sets and some of the early ones I looked at were like the Digital Public Library of America, which was like bringing together a library and museum ah digitized collections from across the country into into one big kind of mega meta databases that was called. ah That was a project that came out of the Harvard harvard Berkman Center. um And that that project actually informed a lot of my questions about big data because I was seeing
03:22.96
Yanni
you know, these enormous collections, but they were aggregated from smaller heterogeneous collections, you know, that each had their own metadata and kind of local um idiosyncrasies.
03:39.30
Jon
Yeah.
03:39.53
Yanni
And, you know, it was really seen as a problem by the originators of that project. Like, how are we going to normalize all these data? So from like a single search bar, you could find anything.
03:50.07
Jon
Yeah.
03:50.34
Yanni
And so, I think pretty early on, you know, I began to ask like, well, what if the heterogeneity is not seen as a problem, you know, it’s kind of an old, old hat, I guess, to say, like, it’s a, it’s not a bug, it’s a feature.
04:05.56
Jon
It’s feature, yeah.
04:06.27
Yanni
um But as I looked into some of the heterogeneous um characteristics of these data, they were so interesting, so human and really spoke about like the places these data came from, whether it was like the local
04:18.34
Jon
Yeah.
04:24.23
Yanni
um regional names for places like, you know, you’d see some records that it was produced in like the Midlands and you’re like, well, where is that?
04:32.08
Jon
Right. Where’s that?
04:33.49
Yanni
ah Or just like the way a date was written.
04:33.59
Jon
Yeah.
04:36.91
Yanni
I did this one project with a student where we, we actually didn’t even look at all the dates across the digital public library of America only um from the New York public libraries subsection.
04:49.21
Jon
Okay.
04:49.28
Yanni
And, and we found, uh, I think it was around 1,800 ways of writing the date in that one sub-collection.
04:57.75
Jon
Wow.
05:00.09
Yanni
And, you know, when you look closely at it, it’s, you know, it makes sense. It’s like sometimes they just have the year. Sometimes they have the year and the month. Sometimes they have the day. Sometimes they’re written in different order.
05:11.19
Yanni
Sometimes they have Roman numerals or like the name of the printer or the
05:13.28
Jon
Yeah.
05:16.28
Yanni
publisher. ah And I began to see these as not just facts about objects, but cultural artifacts. So bringing that back to this issue of big data, you know I said, well, what if we approached all data as being local? and And maybe some of them are more aggregated than others, but they’re all coming from a place and a time from an organization, usually kind of grounded in disciplinary um ways of ways of making data, specific instruments, um and the expectation that they’re being looked at by specific audiences.
05:50.10
Jon
Mmhmm. Mmhmm.
05:58.39
Jon
Yeah.
05:58.60
Yanni
you know not For a lot of these data, let’s say, in in in the libraries, they weren’t expecting that patrons were going to be seeing the data. It was meant for the librarians to use.
06:10.43
Jon
Right.
06:10.67
Yanni
Or some of the collections like an archive is organized very differently from a library.
06:11.26
Jon
Right.
06:16.14
Yanni
And yet, you know, if you want to bring them together.
06:16.81
Jon
Yeah.
06:19.45
Yanni
um and And so that began to kind of inform this larger um this larger inquiry.
06:31.03
Yanni
and And I looked at not just libraries, but I looked at um scientific data from an arboretum also tied to Harvard University.
06:31.53
Jon
Right.
06:41.09
Yanni
I’d been working there previously. And then when I got to Atlanta, I started to look at local Atlanta related things. you know I got really interested in in journalism and how news can be dealt with as data.
06:58.11
Yanni
There are a lot of these kind of news aggregators. you know We’re very used to seeing this now um using things like natural language processing.
07:00.56
Jon
Right.
07:06.40
Yanni
or um housing, housing data. you know That was a big a big part and ended up being the last piece of the book.
07:09.68
Jon
Yeah.
07:14.09
Yanni
um And i and the that chapter really takes on Zillow as this kind of aggregator of housing data and kind of delving into how are the data that go into Zillow’s estimates, how are they created differently in different places and why does that matter?
07:32.00
Jon
Sure. Yeah. Yeah, it’s it’s interesting because I feel like a lot of us get our data, whatever project we’re working on, and we kind of ignore some of these subtleties and you know how they were collected.
07:45.84
Jon
i was just I’m just finishing a project where I’m merging different data sets like at the different at the different county levels.
07:46.17
Yanni
Hmm.
07:52.78
Jon
And just even that, which is supposed to be standardized, um there are different parts of the country that they are not standard.
07:56.69
Yanni
Hmm.
07:59.14
Jon
Or they’re standardized, but they’re standardized in a different way.
07:59.99
Yanni
Hmm.
08:01.42
Jon
like Connecticut has its own thing going on.
08:02.22
Yanni
Yeah.
08:03.79
Jon
and it’s like you know it from From a kind of productivity perspective, it’s frustrating because you have to like carve out this whole other part to like deal with Connecticut and Massachusetts.
08:10.49
Yanni
Hmm.
08:16.83
Jon
But from a local perspective, um it it is the experience of people who are living there.
08:23.12
Yanni
Yeah, yeah, yeah.
08:24.82
Jon
um There is, there is a ah quote in in in your book that I have written down um that I’ve included in a few other things I’ve done and the quote is this and I wanted you to talk ah a bit about it. um In the book you write, what if we learn to see data as situated and partial because of their place attachments, um which which I think is just a great sort of almost summary what you were just talking about, but I was wondering if you talk a little bit more about about how you think about data in that local context and then how that has kind of spurred or encouraged your kind of current line of of research and thinking.
09:00.95
Yanni
Yeah. Yeah. So I think it’s important to say that this idea of place attachment or context and and also you know spatial context mattering or making a difference is not something that comes out of nowhere. I was actually originally trained as an architect.
09:23.00
Yanni
and even though i I left architecture a long time ago, I sometimes describe myself as a renegade architect, I carry with me some of these sensibilities. And I think something that was kind of drilled into me early on was this sense that architecture or good architecture is a a ah response to the context. And whether that context is, you know, the um the site, you know whether it’s how the site is sloped or views that might be available or kind of where the sun will be coming, um where winds are coming from, um access.
09:54.00
Jon
Yeah. Yeah.
10:05.63
Yanni
you know so The idea for me of like thinking about a design as independent of place was a kind of anathema.
10:15.93
Yanni
you know and And as I started to work, you know I was very interested in data and and visualization, as a lot of architects have been.
10:16.64
Jon
Yeah.
10:24.31
Yanni
um But it kind of never sit well with me that this was supposed to be um place agnostic work.
10:35.40
Yanni
and And really, this this idea that place doesn’t matter, goes back it’s kind of been aspirational. It goes back to people like Marshall McLuhan, who talked about electronic media collapsing space and time, or Nicolas Negroponte, who was actually an architect, or isn’t yeah was trained but as an architect, you know wrote
10:49.36
Jon
Yeah.
10:57.17
Yanni
back in the 90s about ah it mattering less and less um where you are when you’re using digital media. And I think this really came together not just as a discomfort but a real problem um with with larger stakes when I started to read work by Anita Chan. um She writes about the problems of what she calls digital universalism.
11:21.94
Yanni
And that’s the notion that it doesn’t matter where you are, who you are, when you are, when you’re using digital media, you’re just a user. um And we’re all the same.
11:33.10
Jon
Hmm.
11:33.42
Yanni
And ah I think in that sense, the stakes are quite high because it means that um you know, there could be potentially an erasure of that context um that could be harmful.
11:48.26
Jon
Right.
11:50.08
Yanni
Some people have talked about this as a kind of digital colonialism or data colonialism, a kind of extractive thing. And and I saw this very much in in some of these library projects where it’s like, okay, we’re going to take your data and then we’re going to remake it.
12:08.03
Yanni
So it conforms to our needs, our expectations rather than asking, you know how did these collections matter you know in the in the context where they were formed?
12:21.88
Yanni
what you know And um how were they tracked?
12:21.82
Jon
Right.
12:25.62
Yanni
How were they organized? And why does that matter for understanding them?
12:31.64
Jon
yeah
12:31.94
Yanni
um And so you know the in the most extreme cases, it can be pretty um pretty daunting. I mean, some of these local um conditions can be, you know, they seem like frivolous, right? But um but other times, you know, it really hits you. I i remember I was having this discussion with a former curator at the Smithsonian, Marta McWhirter, and she said she had done this search in the collection for
13:06.64
Yanni
ah you know, she was she was like, what if I type in black in the Smithsonian, you know, search query bar.
13:14.32
Jon
yeah
13:16.46
Yanni
And she got ah all these artifacts, art objects related to black culture in various ways. oh And then she typed in, she’s like, okay, well, let me search for white.
13:30.71
Yanni
And, you know, of course, white culture is not something that is like tracked and cataloged in a collection like that.
13:37.90
Jon
Yeah.
13:42.15
Yanni
um and And why? And she said, well, because it’s the default.
13:45.74
Jon
Yeah.
13:45.81
Yanni
And so the what she got back was like, you know a painting by someone named white or so of like white mountains or you know objects that were the color white and she said well that’s goes to the very heart of what we mean when we say white supremacy that it is the default condition we assume and then blackness is uh, something that needs to be acknowledged as different from the default.
14:13.87
Yanni
And that’s almost, that’s such an interesting example because it’s not something extra in the data.
14:13.97
Jon
Yeah.
14:19.97
Yanni
It’s actually something that’s missing, but that reveals a lot about the culture in which those data were created.
14:23.11
Jon
Mm hmm.
14:27.81
Jon
Right. Yeah. I mean, the same holds true. You see this in lots of places, right? The same holds true in data on, say, sexual orientation.
14:34.32
Yanni
Hmm.
14:34.49
Jon
I mean, a lot of the ways that data on sexual orientation are collected is, like, you know, one option is, like, straight. And then the next option will be gay, comma, not straight, where it’s like it’s defined as, you know, the opposite of whatever that norm or default is in the front.
14:46.03
Yanni
Yeah. Hmm. Yeah. Yeah.
14:52.11
Jon
Yeah. um So with all this in mind, your work now, um seems to be more about like actually getting into ah communities and working with people.
15:04.70
Yanni
Hmm.
15:06.98
Jon
um We first started talking a few months ago about your um your Map Room project. um Maybe that’s a good starting point. um And we can talk a little bit about you know what that project is and then and then how you’ve engaged people ah to work uh you know sort of I guess very simply to create a data visualization.
15:28.54
Jon
I mean they may not know that they’re creating a data visualization but that’s what they are doing.
15:29.80
Yanni
Yeah Mmm,
15:32.22
Jon
Um so so yeah so maybe we could start there and and or or maybe we start here like talk a little bit about your lab at at university and then we can talk about some specific projects.
15:41.24
Yanni
yeah Yeah, yeah so uh I’ve had a lab for for quite a while called the local data design lab and actually I had a number of wonderful students who, um you know, pass through there and and help me with pieces of the book.
16:03.33
Yanni
people like Peter Pollack and his brother, Chris Pollack, actually two great brothers. and um um But more recently, I have really kind of reoriented my work towards this larger collaborative center that I helped to found in the last couple of years is called the Interdisciplinary um Media Arts Center.
16:25.98
Yanni
And the idea is to really bring people together from across Georgia Tech from different disciplines in order to think about creative approaches to research, um ways of bringing art and design, making um expressive activities into projects, into serious scientific and technical projects as a way of broadening their accessibility, um doing community engagement work, um and just
17:03.12
Yanni
ah putting whatever scientific problems there are in this in this broader material you know social and material context.
17:10.89
Jon
Yeah.
17:11.82
Yanni
And I think one of the one of the first successes we had, actually before the center, but in a way inspired the center, was using the map room in a huge um project about sea level rise in Savannah, Georgia.
17:27.33
Yanni
And in Savannah, Georgia,
17:32.57
Yanni
as many people know, they get hit by a lot of hurricanes um and it’s already a very low lying area. And if you know a hurricane coincides with a king tide, you know there’s often substantial flooding and and it’s projected that you know depending on sea level rise, much of the city could be underwater in the next 100 years.
17:57.72
Jon
in Yeah.
17:59.73
Yanni
And scientists and engineers at Georgia Tech have been working for a couple of years to try and ah record some of these changes ah by putting sensors on things like piers and bridges, people like Russ Clark, and um and um um a host of of scientists, Kim Cobb and and others, ah who’s a climate scientist.
18:26.64
Yanni
Anyway, they so they had all this data. And in in a way, it was like meant to be very local.
18:30.19
Jon
e Yeah.
18:32.98
Yanni
Because here is, you know we think of climate change as being like a global problem, but here you know, these people were collecting data on like, where is the sea level at this pier, at this bridge, and how does it affect this local community?
18:46.40
Jon
Yeah. Yeah.
18:50.74
Yanni
And um they were having a trouble, though, engaging the community in thinking about these problems. And, you know, they had some of the usual kind of, Oh, we’re going to have like ah a kind of open house, or we’re going to have a public meeting and we’ll talk about the data and why it matters, or we’re going to build a data dashboard where anyone can go look at the data.
19:07.54
Jon
Yeah. Yeah. Yeah.
19:15.80
Yanni
And my sense is that it was somewhat unsatisfied with the response they’d gotten. And so I. introduce them to this map room technology I’ve been developing actually with ah with Jer Thorpe originally.
19:31.47
Yanni
Thorpe is a data artist who works out of New York City. He used to run a studio called OCR, um the Office for Creative Research, and
19:38.69
Jon
Okay.
19:41.89
Yanni
He led a project called the st. Louis map room in 2017 and so a lot of the original framing and kind of some of the tech ideas came from that and you know, I basically reached out to him and said ah You know, is this something you’d want to try in other cities and could this become a research project anyway, so we you know
20:03.45
Jon
Mm hmm.
20:07.38
Yanni
brought this to to Savannah and people were immediately and enthralled with it. I mean, it was startling to me. I think kids, you know we took it to a local high school, kids were drawing. you know So the idea with the map room technology is, it’s quite simple actually. It’s a a overhead projector that ah that’s driven by software we’ve written using open street maps.
20:35.78
Yanni
and it projects like the street grid onto a table and then you can lay down paper or canvas and you have this the projection guides you in making a map making as you say a data visualization and um
20:50.94
Jon
Right.
20:54.41
Yanni
the the The fun thing is so the projection operates kind of as a, as I said, guideline or suggestion, but then people have to draw. People follow the lines of the projection and draw the parts of the city they want to draw that are meaningful to them.
21:09.97
Yanni
And because their drawings have followed these lines
21:10.01
Jon
Right.
21:15.12
Yanni
from open street maps, um and they’re geographically precise to a certain extent, then you can overlay any data, existing data you want on top of that.
21:25.67
Yanni
So you can say, OK, you’ve drawn your route to school.
21:26.21
Jon
Yeah.
21:29.03
Yanni
Let’s look at you know where flooding has happened in the past 20 years.
21:30.49
Jon
Hmm.
21:35.74
Jon
Right.
21:35.87
Yanni
um And so that creates a dialogue between their own experience and what are recorded in the data. And I think what’s really exciting about that is It’s a kind of two way thing where on one hand, there may be stuff that’s invisible to them, you know, look they they may say, Oh, I didn’t realize, you know, there had been so much flooding maybe they just moved to Savannah last year and they’re like I didn’t realize this street flooded, you know, every couple of years, um or they may.
22:05.35
Jon
Right. Yeah.
22:09.32
Yanni
see an error in the data because they actually know quite a bit about the place they live and they know things that aren’t in the data.
22:11.79
Jon
e Right.
22:16.32
Yanni
So they may say, actually, this is wrong, or the flooding kind of extended beyond this, or I live right here in my basement flooded.
22:23.27
Jon
ah
22:26.17
Yanni
And um because you know that flooding map you know was yeah created using certain kinds of indicators and so forth, and it’s going to be partial, as I said, and limited.
22:37.57
Jon
Right.
22:40.27
Jon
Yeah.
22:40.54
Yanni
um And then what I find is, in you know as people start making these maps, and they’re huge maps, you know one version, they were like 10 feet by 10 feet. Other versions, they’ve been longer and thinner.
22:52.53
Yanni
ah but we you know we’re all around a table and we start talking and some of the most interesting stuff that happens is not on the map but what passes between people and how they talk about the place they live and its past and also what they they hope for in the future.
23:05.98
Jon
Yeah.
23:13.24
Yanni
And it became a very concrete way of taking the principles of all data or local the book and putting them into practice saying it matters where where data are made, but also where you are when you’re using data.
23:13.29
Jon
Right.
23:28.81
Jon
Yeah.
23:29.05
Yanni
um And Yeah, I could talk a lot more about it, but it was it it was ah it was a wonderful project. It’s still ongoing. We ended up getting a big grant from National Science Foundation through the Civic Innovation Challenge together with um people like Alan Hyde, who’s a sociologist at Georgia Tech and um the ah you know a couple of other um people from around tech. And we also collaborated with um um a university in in Savannah and and the Savannah Public School System. And so it was a very rich opportunity to kind of see how this creative activity could
24:14.89
Yanni
really add to in a substantive way, add to this scientific project in a way that and NSF could get.
24:20.96
Jon
Right, right, right. But it it also strikes me that it empowers people with their own data, their own experiences in ways that maybe be your sort of standard ah community meeting, even if they’re allowed to get up and share ah their story at a lectern or something like that.
24:39.61
Yanni
Yeah.
24:42.85
Yanni
yeah
24:43.57
Jon
But but the the act of drawing and the act of standing around the map empowers them in in ways that some of these other ways just maybe they just fundamentally don’t.
24:55.55
Yanni
Yeah, yeah. I think one of the most interesting things is it allows people to disagree about their representations of the city. One of the most wonderful things about it is because you have multiple people working on the map at the same time.
25:12.85
Yanni
they might not necessarily be in alignment with how they’re drawing. And these, I remember these two kids, middle school age kids at a Savannah public school, where it started to draw the Savannah River from opposite ends. And on one side, it’s this crystalline blue surface.
25:31.45
Jon
yeah
25:31.81
Yanni
like pristine and in the other side it’s like orange and brown with like debris in it and they meet at the center and there’s really like you know this line at the center where they met and of course they had this discussion about how one kid thought the river was this beautiful natural resource and the other one saw it as heavily polluted and a danger really.
25:44.07
Jon
Great.
25:55.05
Jon
Yeah. Yeah.
25:57.91
Yanni
and And then you know they could ask, OK, well, are there data that exist that could help us? learn about this. And there might not necessarily be the data out there, but what I like about is it’s starting from their experience rather than starting from the data.
26:14.48
Jon
Right.
26:14.93
Yanni
So I always try to get them to start from, what is your experience? And then asking, are there data that speak to this or not? Because I think it’s important for people to be to be able to say, actually, no, I care about this thing that’s not in the data.
26:31.00
Yanni
And maybe that can be the start of a process where they um go out and try to collect some data or encourage the city to collect it or partner with scientists and that actually happened in the sea level sensors project because a number of people would come to the mapping sessions and they would say you know what I’m really worried about remember this one woman was like why is it that when I come out of my house in the morning there’s like white particulate matter all over my car and what is going on with the air quality.
27:03.18
Yanni
And that was not at all what we expected to talk about.
27:05.20
Jon
Yeah.
27:06.39
Yanni
But because of the openness of this system, sometimes I call these systems open data settings.
27:09.92
Jon
Yeah.
27:13.23
Jon
Yeah.
27:13.33
Yanni
you know they’re They’re places in which anything can become data. you know Your story can become data, um and ah rather than this kind of more closed scenario where it’s like, these are the data we have.
27:21.62
Jon
yeah
27:26.95
Yanni
What can we say about them? Anyway, so she started to draw, like, here’s where I’ve also you know smelled you know problematic smells.
27:37.29
Yanni
And you know eventually we realized, OK, there’s like a paper mill that’s nearby, and that’s causing a lot of pollution.
27:41.70
Jon
ah Yeah.
27:44.59
Yanni
um and And then um you know within a year, this Russ Clark and and some other engineers started to get some simple air quality sensors and put them you know in a neighborhood in Savannah.
27:59.01
Yanni
So I think it shows how, yeah, that there is a kind of bidirectionality that opens up.
28:06.07
Jon
Yeah. Yeah, I also wonder about the data that the parts of cities that that participants don’t draw.
28:15.20
Yanni
Hmm.
28:15.19
Jon
like so So you’ve got this projection of some area of the city on this big or small piece of paper.
28:20.51
Yanni
Yeah.
28:20.85
Jon
People are drawing on it. And then have you seen times where people talk about what they haven’t drawn? Because I feel like what’s missing is all is some can be like equally as interesting of what people have drawn.
28:30.93
Yanni
yeah Yeah, yeah, yeah. There’s actually an interesting project going on right now by one of my PhD students, Mohsen Yousufi,
28:46.95
Yanni
that really stemmed from something we heard from a number of middle school age kids in in this Savannah project. And it was, we’d say like, we’d ask them to draw like a place that they frequently go in the city or something like that. And they’d say, well, I don’t really know anything about the city.
29:12.54
Jon
Yeah.
29:12.89
Yanni
Um, and so I can’t, you know, I can’t draw, I don’t know what to draw. And, you know, we struggled with that early on, like quite a bit. It’s like, what is going on here? And, and he has framed it in this interesting way as a, what he calls an epistemic obstacle. Um, um,
29:34.64
Yanni
And because it raises this question, is it really true that they don’t know anything? Or do they think that what they know is not important?
29:43.37
Jon
Right.
29:43.57
Yanni
um Or, yeah, is there something else going on? Maybe they think they they’re not sure if what they know is true. Or, ah yeah, so there are all these other explanations.
29:58.82
Yanni
And I think what’s so interesting for us about the Map Room and its various incarnations is it becomes a place in which to study how people relate to data and some of the ah problems and that come up in limiting them sometimes, you know, from ah engaging with data in certain ways or um becoming fearful.
30:26.82
Yanni
you know One kid said at some point, is like do we really have to talk about this because it’s scary.
30:32.04
Jon
Well, right.
30:32.39
Yanni
you know It’s scary to me to think about my neighborhood being affected by a hurricane or flooding. um And I think it’s important for us to To see that, to hear that, and to really take in that data, for example, are not just rational, logical elements, but they carry this emotional weight.
30:54.67
Jon
Yeah.
30:55.10
Yanni
And I think that really differentiates, let’s say, how humans deal with data and how computers deal with data.
31:01.14
Jon
Yeah, yeah.
31:01.23
Yanni
You know, for humans, data are a felt experience, both like perceptually, you know, we see the data, um it’s a sensory experience, but also we have to manage our perception of the data with our own complex feelings about the subjects they represent.
31:20.67
Yanni
And computers and, you know, we could talk about it in terms of AI, doesn’t have that. And in, you know, one could say, oh, well, that’s the advantage because AI can be impartial.
31:27.31
Jon
You’re right.
31:33.89
Yanni
But, you know, I think those emotions matter because emotions are what motivate us to take action and do things.
31:39.95
Jon
Mhm. Mhm.
31:41.67
Yanni
And so we don’t want people to be just objective and impartial. We want them to be motivated. Obviously, when I want them to be biased,
31:52.95
Yanni
But I think it’s in the map room, you can see that people have a position. They have a standpoint, you know, um in the language of, you know, standpoint of epistemology, we could say, you know, um and and and we can start to see that and how it plays out in the way that different people draw or don’t draw parts of the city.
32:05.80
Jon
Right.
32:14.11
Jon
Right. Right. um I want to close up with one last question. I’m thinking about folks who, you know, who listen to the show who do their day-to-day work, right? They collect some data, they download some data and they analyze the data, they make a graph or a dashboard, whatever they’re doing.
32:34.76
Jon
And I’m curious what you would tell that person who works on big data sets and we can, they could mean anything in this in this in this context, but what you would say or recommend to people to have this local sort of perspective as they’re working on their various projects.
32:41.77
Yanni
Yeah. Yeah.
32:53.87
Yanni
o
32:58.34
Yanni
Yeah. It’s such an important question. the The last line of the book actually goes something like this. ah Don’t take the existence or the accessibility of data as permission to remain at a distance. Use it as a starting point to get closer to, to learn more about the people and the places that the data represent. And i I really think this has been
33:34.43
Yanni
at the root of everything that’s enabled me to write this book and do these projects. At the end of the day, I’m not a domain expert in any of these areas where I use data, but I’m able to do the work because I talk to people who have either created the data or they um they use the data every day and they really know it or they maybe are
33:46.76
Jon
Yeah.
33:58.97
Jon
Mm hmm.
34:03.10
Yanni
in the data, they’re subjects of the data, but they are able to connect me to this broader um system of knowledge.
34:16.17
Jon
Right, right.
34:17.25
Yanni
And I like to really think about, sometimes I say, well, data are like the index of a book, you know, the index of a book is great. It’s really useful for understanding what’s in the book and you know can can tell you so much, but they’re much more useful when you use the index with the book.
34:38.05
Jon
Yeah, yeah, yeah, yeah Yeah
34:38.25
Yanni
And so you know in a way, they become pointers. And I often ask my students to create what I call data guides um before they start working with a data set.
34:50.41
Yanni
And that means you know being intentional about going out and having a discussion with somebody who knows this, data source, um looking at how are other people making arguments or visualizations with the same data? How are these data created differently in other places? um What’s an example of like an anomaly in the data? um And so they are starting from a more engaged position. And it is very hard for them, frankly. I just went through this with some students. Often they come back initially and, you know,
35:28.39
Yanni
they were embarrassed to call someone or they called somebody or reached out over email and they didn’t hear back. So they gave up, you know, a lot of people who work with data, um I mean, a lot of us in general these days, you know, turn to data because it seems easier than dealing with the messiness of, can I trust this person and what do they know and not know, or what is their perspective?
35:45.26
Jon
Yeah.
35:50.52
Jon
Right.
35:53.80
Yanni
But I think it would be a mistake just from all the different data sources I’ve worked with. As soon as I talk to somebody about it, it opens up a whole new set of questions and often I was looking at the wrong thing or I was misinterpreting something.
36:15.97
Yanni
um So it is really powerful ah to have to build those relationships and see data work, whether it’s building visualizations or analysis as a social process and not as a either automated process or or something you do on your own um in the sense of like mining you know the data.
36:27.51
Jon
isn Right.
36:38.28
Jon
Yeah.
36:39.21
Yanni
Because the answers aren’t all in there.
36:39.27
Jon
Yeah.
36:41.42
Yanni
The answers, there may be pointers to great answers, but there data are always incomplete, in a sense.
36:50.70
Jon
Yeah.
36:50.89
Yanni
They’re limited because ah you know the whole idea of creating data is to efficiently represent a very complex phenomenon.
37:03.11
Yanni
so you know, that’s the advantage that they are smaller than the original phenomenon, but that comes with certain problems.
37:11.98
Jon
Yes, right.
37:12.49
Yanni
And I think most good scientists and researchers know this. The problem is when we have open data and a much broader set of people, including people who have nothing to do with the domain or have not been involved in collecting the data and so forth,
37:32.19
Yanni
expect that they can use the data without any of that context that you run into a problem.
37:39.90
Jon
Yeah. I mean, I think part of the problem is, as you as you mentioned with some of your students, like this apprehension, fear, what i have you, of talking to people, of reaching out to people and figuring out how to do that and you know having those conversations is is is hard for lots of people, especially for people who are are more comfortable and like to sit behind the computer and you know write the code.
38:02.19
Yanni
Hmm. Yeah.
38:03.76
Jon
and yeah
38:04.89
Yanni
Well, hopefully we’re modeling for people why it’s so great to have discussions.
38:05.07
Jon
um
38:08.01
Jon
Yeah, yeah, exactly, exactly. Yanni, thanks so much for coming on the show. So the book is all data are local.
38:13.17
Yanni
Thank you.
38:15.20
Jon
um ah The project we talked about is the is the the mapping project, but there’s lots of other projects on your site. And I’ll make sure that to put the the link on the on the show notes.
38:25.03
Yanni
And I should note mention the book is open access, so you can get it free online.
38:28.13
Jon
Right. Yeah, or you get the lovely hardcover version, which I have right here.
38:32.92
Yanni
Yes, it is nice.
38:33.76
Jon
So Yanni, thanks so much for coming on the show. There’s a lot of fun.
38:37.57
Yanni
Thank you.
