Ep 02 – The Constitute Project Pt. 2

[0:00:04 Speaker 1] Madison have been able to draft a grader constitution with access to the Internet instead of leafing through hundreds of leather bound books. Welcome back to the connector, where we bring together innovative, groundbreaking and collaborative research inside the U. T. Austin political science universe. I’m your host, Stewart Tendler. This episode is Part two of our three part series on Constitute the World’s Constitutions to read, Search and compare. If you missed Episode one, go back and check it out before diving into Part two. In this episode, we continue our discussion with Zach Elkins and Robert Schaeffer about constitutes work in the T A. C C visualization lab and the role of visualizations in generating research. Can simply arranging data points in a certain way on a wheel or a map or a timeline? Tease out research questions we wouldn’t otherwise notice? Is there an optimal way to arrange data points that sparks public imagination and learning rather than boredom or confusion? Zach and Robert also consider how the constitute project maybe an active agent of the very topic that they’re researching. Diffusion. How do ideas spread out globally across nations and continents? Two researchers have a moral responsibility to control diffusion of their information. For example, is it morally acceptable to provide constitutional drafters with repressive constitutions? The

[0:01:41 Speaker 0] purpose behind these visualizations, if I heard right, was to kind of help generate research questions. So I’m wondering what kinds of research questions of any these visualizations have generated and sort of where things stand with that

[0:01:56 Speaker 3] we have this visualization, which in some ways required us to collapse rights within certain categories just as a way to manage this physical space. You had to put rights and categories, and that’s a major conceptual question unto itself. How Shwe categories rights is gonna sort of somehow be based on their ah, substantive content. May we choose political, civil and economic and social rights? Would we group them based on when they evolved and talk about certain generations of rights? Could we talk maybe about sort of positive versus negative rights and that sort of thing? And so that’s something that social scientists have been struggling with for a long time, how to think about rights and really other political institutions and how to conceptualize them. And I think putting these on ah, sometimes small screen forces you to make those decisions in a way that you wouldn’t otherwise. And one thing we did and are still doing with that one. Visualization is allow people to conceptualize writes in different ways. So we set it up so that you can choose to look at rights based on whether they’re political, civil, social, economic rights. But you can categorize them. Ah, in really an infinite number of other ways in the way we set it up and give you these phenomena differently. So that’s one set of research questions, which is mostly sort of more of a conceptual set of questions. But that’s opened up our eyes to different ways of viewing these things,

[0:03:27 Speaker 4] some exact. And I have both, you know, one of our driving interests, and it sort of constitutions in general, is is the diffusion of ideas. So how how ideas spread around the world and the patterns by which they spread, you know, did a spread geographically or culturally or which pathways they spread in. And these sorts visualizations air are often a great way to sort of see that spread in action on. And it helps you sort of think about how that sort of pattern of sort of spread of ideas like right, some things that might be happening. This wheel of rights visualization was a great way to see that part of this wheel was, ah, sort of time, dimension and a geographic dimensions. He could see you know which rights became popular, particular points in time or things like that. It also helped us think about sort of other ways that we might see this kind of diffusion in action. So, um, in the social sciences these days, something people love to dio is look a plagiarism and not plagiarism in the in the sense that you might think about in the academic sense but the sort of re use of tax, the copying of text from one place to another in often laws but in in sort of other kinds of speeches and, you know, political dialogue as well. And so constitutions are a great place where you can look for this sort of text re use. And again, this is another place where the the visualization lab stuff was helpful. We could sweded click and see the text of certain constitutional provisions and see sort of re use of of language over time and that was something. It’s a lot of fun. It’s an exact and I are actually have sort of ongoing work on Is looking for this sort of this kind of clustering and in text usage patterns, things like that.

[0:05:03 Speaker 3] Robert I just wrote a paper and went to try to validate different ways of assessing textual similarity. And part of that came about because Robert especially is struggling with this really difficult problem of how you assess the similarity of text, which turns out to be really difficult problem. And it became, you know, in such stark relief, um, something that we wanted to look at Onley because when we saw museumgoers play with our interactive often well caught their attention and resonated Ah, in some ways most loudly was this data point that we put up there, which is a 21 point our visualization. This is at the museum. You can choose to countries and look at their text on any given right side by side. And then Robert over laid that with a stat that showed how similar the two texts were, and people would just puzzle over that like, Okay, that’s 29% similar. But why? Um and Robert had, you know, different algorithms for calculating that, and it was clear that you’re getting different answers, you know, as we modified as he a modified ah, calculation. And the director of the museum was especially perplexed by this. And so he had this one excerpt which he always went to, which is that, um he was focused on the Fourth Amendment in the U. S. Constitution and wanted to see the version of that in the Japanese constitution, which happens to be very similar. Um, and he would always sort of look at the percent similarity. And there was this kind of reference case for that, Um, but that led to this this paper we wrote which I think in many ways will be one of the more significant technical papers to come out of this project. But it’s a really interesting problem, but one more idea. I wanted to mention, um, which came up when Robert was talking. Is this I idea of diffusion, which was one of the major research ideas that lead to the creation of the Comparative Constitutions Project. But one interesting thing about constitute is that it’s allowing us in some ways to intervene into the into this problem of the diffusion of ideas. Ah, because, of course, the research question behind that is whose constitutions air similar to whose and sort of a long if you want which networks. So is it it, Ah, countries that share colonial ties, countries that are in the same region, countries to speak the language that have the most similar constitutions or what? This is a problem that many people puzzle over over a different phenomenon. But it’s really interesting to look at constitutions in this sense, because they’re famously traded back and forth across countries and across time and, um, in putting constitute forward. We have such a wanted to intervene in the constitutional writing process and intervene in a way in which we put all constitutions in front of people such that they wouldn’t just be cherry picking certain constitutions, say the most famous ones or the ones the most successful countries. They would have them all. And so, in so doing that would change the exchange of ideas across countries and really hopelessly metal in our own research question, in kind of an interesting way

[0:08:17 Speaker 0] can we talk a little bit more about that because that, I mean, whether it’s sort of stating the obvious or overreaching or some aspect of both through, you’ve kind of become an agent of the subject, your study, Yeah, I mean, I don’t know. I kind of

[0:08:32 Speaker 1] it’s really intriguing

[0:08:33 Speaker 3] question. In some ways, um, I think of it about this the same way anthropologists think about participant observation in that when they’re involved in understanding subjects and actors in how they work in the natural world. Of course, anthropologists recognize that they themselves absolute as observers are having an impact on what’s going on. Um, and I think, you know, obviously, anthropologists try to adjust in some ways for that. Um, you know, we’re not. We’re humble enough to think that we’re not having that much of an impact on the world, although it is the case that a lot of people are using, constitute. And I think it’s fair to say that anybody working on a Constitution today and the drafting of the Constitution is probably going to constitute, so it’s likely that we’re having an impact. Um, and for the most part, I like that idea. Since I think our impact is a positive one. That is, what we’re doing is putting in front of drafters of representatives of texts and, ah, that should be a good thing. Assuming that they can take that set, um, and come up with an idea that maximizes whatever objective they have in a better way than they would if they only had, ah, highly sampled set of text.

[0:09:55 Speaker 4] Sometimes the impact is is unexpected to we’ve we’ve done some sort of impact analyses on on constitute looking at our usage data from various places around the world. One good example of the sort of unexpected effects is if you look at that, the rankings of the most access to constitutions on our site that most of the top five or 10 or what you would expect, you know, the U. S. Is always the first is always the top. A few others, like France and Britain and so forth are are high up there. But one that is always in the top five is Afghanistan. Why Afghanistan? Afghanistan’s first alphabetically, and people always, you know, go to the 1st 1 This sort of learned how the site works and and take a look and sort of get an idea of what’s out there. So we’ve sort of jokingly started talking about the Afghanistan effect. And, you know, maybe in 10 years or 20 years, we’ll start seeing this, you know, Spike in constant that looked like the Afghani one. Who knows? But it’s sort of thing that you know, might happen.

[0:10:54 Speaker 3] Yeah, And so, in some ways, in terms of social engineering Ah, one, I think implication of that is if, as we’re doing version 4.0, the site, we would sort these countries differently. Um, And what happens when we do sort thes? Who knows? Maybe then we’ll get a more representative set. But, um, we like the Afghanistan effect as kind of a social science e thing. But in the end, it’s one of those things that make us wonder about whether our factors on the whole positive. But, uh,

[0:11:24 Speaker 0] do you think again, I don’t know if I’m getting the phrasing correct here, but you paint a picture of sort of, ah value free intervention right where you’re just providing more information. There’s no nothing attached to that information. Do you think that it’s being received that way. Or do you have any reason to suspect that somehow, um, access to these documents or the site itself is in somehow influencing the drafters to say no sample? A. Is better than sample be? Or do you have any fears of that?

[0:12:05 Speaker 3] It’s an interesting question. I think. On the whole, what we’re doing is just providing more information and an exhaustive set of information. And these drafters are, for the most part, smart people. And they have their own biases, and they’re going to use them in their own way to advance whatever political interest they have. Ah, so I don’t think there’s too much weaken due to sway them. In fact, I think a lot of what goes on is maybe they’re using constitute, as a rhetorical vice toe argue for whatever they wanna do to begin with. One thing we struggled with is whether we should put all countries or just all quote unquote Democratic countries, assuming we can identify that set. So there is a question. Should we be putting North Korea’s constitution up there? Um, is that something that we want people using and learning from? I mean, I think on the whole. Yes, only because I think most users are smart enough to select which countries they want to learn from, and they’re going to do that anyway. So I don’t think there’s much of a concern. And historically I think there’s a lot to be said, Ah, for seeing what somebody’s closed countries were doing. And some of these authoritarian countries air doing certainly is political scientists and other scholars. We want to know what they’re doing, so I I would never want to repress that information. But it is true. I mean, we have to think about who’s using it and under what context. If somebody ah was hard pressed to write a Constitution in three or four hours, I wouldn’t put the world’s constitutions at their fingertips. I might curate that list a little bit more, and we’re gonna put out I think, a print version of this which would hopefully go on the conference table of any drafting commission in the world on that print edition on Lee from for space reasons, probably won’t include the constitutions of some of the more repressive regimes in the world.

[0:13:56 Speaker 0] If we could return to sort of the diffusion issue if you could provide us any additional context, the bigger subject sort of seems to be the role of technology in diffusion. If you could just talk at all about where technology fits into diffusion historically and just in your theoretical thinking

[0:14:20 Speaker 3] of you know this subject well, it’s a really interesting topic. I think it could be a radio show unto itself. Um, but I tend to think that, um, diffusion is just sort of a very basic part of anybody’s behaviour, regardless of what they’re doing and that there’s very few things that are very original in the world today. Whether it’s an architect putting together a building or an artist, Ah, painting a painting or somebody drafting a piece of text, I think they’re always at least ah, subconsciously modeling what they’re doing on prior ideas and there. I think there’s very little that’s truly original in this world, and so I think once you recognize that that’s a basic part of somebody’s repertoire, then the question is, how can we best assist somebody to do those things? And so if somebody were an architect and was designing a building, um, for a certain group for certain activity. Ideally, what we want to do is put in that architects hands, large set of models. Ah, that and blueprints that might be helpful. And in many ways, I think that’s what’s going on with technology. If you think about what’s going on in the last 200 years, that wasn’t really possible before the information age. And certainly people traded books and prints throughout the ages. And if you think about Madison, what he was doing when he sat down to write his draft of the US Constitution, he had a trunk load of books that his buddy, Thomas Jefferson had sent him and which Madison took from his own library. And he sat down in Philadelphia and spent, um, what had to be, Ah, really interesting set of weeks working his way through those books. So there isn’t that much. Probably that’s different from that exercise. Compared with, say, the exercise of somebody looking through whatever is available online, except for the fact that it’s much more efficient now and maybe more enjoyable. I don’t know we did this experiment, which is kind of interesting, which we allowed people to write a constitution with our website, which we love ah constitute, but then also allowed people to use whatever sources they wanted Teoh and one of my favorite results from that experiment, which admittedly hasn’t been done many times. And so this is a small sample. But one thing they said was they enjoyed the experience a lot more when they’re using, constitute. And there’s something about being ableto access, exactly the right information right away and move from the research to your writing really quickly, which is just much more pleasurable. And so I think that’s one thing that’s going on with technology. Of course, we could argue about information overload, etcetera. And what what you can do when you have to drink from the fire hose of information we need. Teoh need to actually sit down and do something that could be overwhelming and maybe not as enjoyable. But I feel like we’ve built this sort of more curated experience, which makes it fun.

[0:17:25 Speaker 4] One thing that maybe is a little bit different as faras, the sort of diffusion via constitute versus diffusion. The other means is we mentioned that we sort of we hope that we’re not kind of directing the process as much as it was directed in the past. Eso Nothing in Life is new consulting on the sort of creation of constitutions and new. Either There’s sort of a long history in a a real cottage industry of people, especially in the last couple of decades who have consulted on the creation of constitutions. Eso, after the collapse of the Soviet Union and a number of this sort of Soviet states in Eastern Europe were looking to write constitutions, and there was this sort of flourishing of constitutional activity and and there were a number of sort of American lawyers and law school professors and so forth who assisted in that process. And one of the things that we sort of ended up in that process is to be end up with a a fairly kind of thematically coherent set of constitutions in that in that area, you know, there’s a lot of regional variation, but there are also a lot of similarities, which sort of makes sense, right. These were people who are writing at the same time and who were working with the same set of consultants to help sort of Veldt, thes thes documents and with sort of tools like constitute online platforms that sort of provide this common knowledge fares. There’s less a margin 80. And so we you know, again, it will be interesting to look in the future and see if you know we see less sort of temporally coherent constitutions. If you know if if that sort of effect sorts to it starts to go away a little bit. That’ll be fun to see. For now, it’s just sort of something to inspect. It weighed about, but it is a potential difference, with sort of the technology driven consult thing versus sort of human curated, human driven consulting. That was sort of the way it was always done in the past.

[0:19:11 Speaker 1] That’s it for Episode two. I hope you are enjoying this conversation about constitutions, diffusion and social signs. If you haven’t already, check out the constitute website at Constitute project dot org’s and please join us next time for the final installment of our discussion. As always, to send feedback, you can find me at Gov dot utexas dot edu