Ep 03 – The Constitute Project Pt. 3

[0:00:06 Speaker 2] constitutions have a heritage lineage you can trace back in time, not unlike a family tree. 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. Thank you for joining us for the third of three episodes exploring the Constitute Project with Zach Elkins and Robert Schaeffer. If you missed either the 1st 2 episodes, do yourself a favor and come back after you’ve had a chance to listen. In this final installment, we turn to a discussion of every social scientists to favorite thing data. Zach talks about creating machine readable data, Google knowledge graphs and being a political science data pioneer. Then we look to the future and explore how collaborative technology is working to improve the process and quality of group writing. Robert and Zach talk about the rewards that come with the outreach component of their work, and we close with a look at the Spanish version of Constitute way.

[0:01:17 Speaker 4] Ah, well, let’s talk about data. The project is generating a lot of data collecting a lot of data. This is a big data project.

[0:01:27 Speaker 2] One thing

[0:01:28 Speaker 0] about the data behind constitute is that it’s in a very new and different form. So it makes use of data from the coveted Comparative Constitutions Project, which is very traditional data, um, in in some ways kind of a matrix form. But what’s going on with constitutes that we wanted to put out this data that would be machine readable, the machines confined and hit news. And for that, we wanted to use this new form of data that’s coming out and sometimes known as Web 3.0, Ah, The idea behind the state, uh, is that every single data point, whether it’s a Constitution in a country or anything in our data set will have its own unique identify are out there on the Web, and as such, it could be connected to anything else on the Web. And so for those listeners you familiar with what’s going on. And this is now 2015 with Google’s knowledge graph, which is taking all this data on the Web and trying to harness it and trying to put together some cards that come up on a Google search. This is exactly the kind of thing that’s now doable with the kind of data we have. I think we’re probably the first political science data set to put our data out in this form, and it’s recently been used in Google’s knowledge graph. So if you look up U. S Constitution and Google surge, what will show up is this card at the very top of your search results, which has the U. S. Constitution broken down its various components. Um and so you can in that ah form read the US Constitution, which is in some ways generated from the data on our site. So Google is just dragging the data from our site on the U. S. Constitution, which has been broken down its various components because of the python code that that Robert wrote. And now they’re in some ways just showing and and surfacing that data on the U. S. Constitution right there at the top of your search results. And it’s also the case. And I think the coolest aspect of what’s called the semantic Web is that all of our conceptual categories, all the ways that we carve out and think about constitutions, you know, that’s just our way of thinking about constitutions inside our own head. But of course, other people have other ways to think about constitutions, and we could take our concepts. And it’s we have this data set just essentially conceptual data set of how we think about our constitutions, that we could marry that with other conceptualization. And if somebody wanted to search across our data using their own concepts, they could. They would just have to marry their concepts with ours. It’s somewhat complicated, but I think one way to think about it is Eflin A s and other people who 150 years ago we’re carving up the natural world into their own conceptual categories. And, ah, class order, violent, etcetera. Um, that’s just one way of thinking about the natural world. But of course, competitors that other ways of thinking about the natural world and imagine if they could marry their different ways of different sort of organizational schema and ah, and show these organisms in the world as categorized in different ways. It would be really cool, and we wouldn’t necessarily be left with this highly standardized way of thinking about ah natural things and and worrying about whether things like an olive should be a fruit or vegetable.

[0:04:53 Speaker 3] That analogy to biology. It’s not just analogy. We’ve actually done a little bit of work on that, mostly for fun. But we have. We’ve tried to build a constitutional tree of life of sorts. It was mostly sort of, ah, exercise out of curiosity. But we, uh we use the same data set, this same sort of semantic Web data set to build a what’s called a final genetic tree of constitutions so showing mostly statistically which constitutions were sort of related to one another, which could be sort of viewed as his ancestors of one another. Kind of conceptually made some nice sort of tree tree of life type visualizations of constitutions based on that again, something sort of mostly for fun. But it was at another neat way to sort of to view the data and to try to induce sort of research questions and stuff like that to Tim Urge out

[0:05:43 Speaker 2] before we break.

[0:05:44 Speaker 0] And I talked about the collaborative drafting stuff.

[0:05:46 Speaker 4] Yeah, you’ve integrated Google docks into the process, and people are now using Google docks to draft Constitution’s

[0:05:54 Speaker 0] Yeah, but why don’t you tell us about

[0:05:55 Speaker 2] it? It’s one of

[0:05:57 Speaker 0] the most interesting things about the website is that having partnered with Google weaken now, in some ways, integrate seamlessly is with some of the other products. And Google docks, I think, has become the industry standard for collaborative drafting software. And so, if you want to write something with somebody, probably used Google docks to do it now. And since constitutions should be a collaborative draft to document, it only makes sense that some of that might happen on Google docks. And so we’ve integrated Google docks nicely with constitutes such that once you find your excerpts that you want to analyze, you can export them directly into Google docks, and that set of excerpts could be the basis of discussion for any set of collaborators you have going forward. But one thing that we’re doing, which I think is, um, even more promising is to address directly some of the problems in the some some of the pathologies of group writing. And we all know that because we’ve all drafted things together with co authors, and we know what happens when you draft some of the co authors. There’s all sorts of ah inconsistencies, whether it be redundancies or just lack of harmony throughout the document, or really sort of long, lengthy documents that you know which have parts that don’t speak well with one another. And that’s a really hard thing, and it’s hard to avoid that. But I think it’s even harder to avoid that when you’re drafting with people that are, in some sense adversaries. And so when you’re writing constitutions with other state legislators, ah, from another party who don’t even share the same objectives that you do. All those issues and all those pathologies of group writing are exacerbated. And one thing that we don’t have when we draft together is any sort of nice set of devices to a decision making. And so, if you’re a group of authors were struggling over some text, yeah, you might comment in the margin. And currently on Google docks, you can reply to those comments and you could have a nice I met a discussion about it. Ah, but one thing you can’t do is to make decisions in the end, and it’s hard to make decisions about text, but we’re developing right now. We’re working with Google docks to develop Samat on technology that’ll allow drafters tax to make decisions on a dock. And so it might be a simple vote in which people make a proposal and to a thumbs up, thumbs down on that proposal or maybe some survey questions. They’re even more nuanced, say their opinion above some maybe orginal scale on some proposal, Um, or even maybe something that’s not even a majoritarian way of making decisions. But it’s more consensual. That’s often what happens on Wikipedia is that decisions are made on a consensus basis and not by voting. And so we’re working on integrating some of the software from Wikipedia into a Google doc so people can write constitutions on their favorite piece of software but also make decisions about text in a collaborative way. And that’s direction we’re going right now. And I hope this spring has something to show for it.

[0:09:12 Speaker 4] You know, the term kind of cutting edge. It gets thrown around a lot and everything is he. Oh, this is such cutting edge work and this or that right? But not only are you kind of involved in this very collaborative research project and researching a process that is collaborative, but you’re changing the way collaboration happens to write. And with Google, I mean, I think that’s another aspect to it, right? I think there’s this Ah idea that there’s always a trade off between kind of, you know, sticking to hard social science and, ah, being, ah, diverse and methodologies and being diverse in approaches to questions and being kind of intellectually and academically broad. You all don’t seem to be sacrificing in any way kind of very the mainstream social science aspect to this research into this project. And yet you’re involved in all these different things and and really bringing in different fields. Whether it’s come sort of the computer programming, statistics, stuffs and biology and really incredible. I think

[0:10:26 Speaker 3] it is anything I’ve learned from from Zach and from everybody else it in this department. It’s that you know, being brought and being sort of open toe ideas and inspiration from a lot of different fields is, you know, it is the way that sort of good work happens. There’s been heavy statistics in computer science. There’s been sort of serious philosophy and, you know, philosophy of science, philosophy, of sort of government and all that that that’s gone into this this project and it’s ah, I opening it was how I got my start and is being an academic. This is working on this project. I think sort of the philosophy behind it has really become what’s guided me in my in my work ever since.

[0:11:05 Speaker 0] I mean, that’s one thing that’s been kind of gratifying is that in some ways, um, engaging with the public and engaging with broader ideas hasn’t meant that we had to sacrifice some of the research ideas. In fact, it seems like every time we do an outreach project, there’s an element of it. Ah, which becomes whether it be a technical research question or more substantive research question something that we integrate in. Um, in a political piece of political science research, you know, these disciplinary boundaries air breaking down. The political scientists are adjusting and using, you know, ideas and methods, um, from all over the place in a really nice way. But, uh oh, can I say one thing about Spanish and Arabic They used to be when we launched in December 2000 now, September 2013 it was only in English, and then we realized that you know, people want to read constitutional text in their native language, and we also realized that conditionally least a really active part of the world is the Arabic speaking world. And so we decided to launch in Arabic in December 2014 which had a subset of texts in Arabic, and the rest of the site was in Arabicas Well, and were somewhat curious what the effect would be. And we had a sense that theoretically, people should be reading these text their native language. I mean, it’s hard enough to read constitutions in English and interpret what they being. It’s it’s hard enough to do it if it’s not your native language, Uhm so we put these things out in Arabic and then the question was okay, what’s going on? If we were to look at the activity online in Arabic speaking countries, are people behaving different are differently? Are the interacting with these text differently? And we’ve just begun to analyze those data. But it does seem that people in predominantly speaking Arabic are spending more time on the site, and we’re getting more visitors. Um, and he’s also been able to identify Arabic speaking individuals based on the browser settings on their computer, and those people are also spending more time on the site. So it seems like what we suspect it is happening. And now we vowed to launch in Spanish, um and ah which is a really important country only because a lot of people speak it. And people in the Spanish speaking world, Ah for the most part aren’t is converse in English as our individuals and other countries, mostly because their incentives for speaking English or much lower, because there’s a critical mass of Spanish speakers. And so, if you want to actually engage in constitutional discussion in Chile, where ah will be writing a new constitution for that two or three years, you really have to engage in Spanish and I have text that are in Spanish. And so now we’re looking to launch constitute in Spanish in September of 2016. We hope in time to allow the chill lands and others toe actually modify their constitutions and using constitute

[0:14:14 Speaker 4] Have you been in any discussions with people there and have you said this is coming in there they know and are

[0:14:22 Speaker 0] Recently we’ve partnered with some colleagues at the University of Lausanne days in Bogota, which is a great place toe study to analyze constitutions for a variety of reasons. But we’ve decided to partner with them, among others, to help translate our content into Spanish and, ah, help ah interact with scholars and constitutional drafters in Latin America. And so last September we participated in a week long school for social and economic rights in Bogota, at theater syphilis on this and with about 50 judges and activists and scholars started planting the idea of a Spanish constitute and started to planet, and we’ll be working with them to bring this to fruition. And I think it’s gonna be received warmly. And we conducted mostly activities last September in Spanish and, ah, people are really receptive toe having this kind of information in Spanish. And I should say also, I mean, you know, Texas is the ideal place to do this, because our ensued for Latin American studies is one of the best in the world. And, you know, right here in our department, we have mawr, Latin American ists than you know. Most of other departments have comparative ists. I think we have about nine Latin American. It’s just in our department. So we were well suited to really bring this to fruition.

[0:15:52 Speaker 2] Well, that brings our three part discussion of Constitute to a close. I want to thank Zach Elkins and Robert Schaeffer for bringing us to the cutting edge of constitutional research. If you have not yet been there, the site is constitute project dot org’s look for the coming Spanish version, and if you find yourself in Philadelphia, stop by the National Constitution Center and find out for yourself just how fun constitutions Ca NBI to send feedback, you can find me a Gov dot utexas dot edu