216. Digitally Deconstructing the Constitution with Dr. Nicholas Cole

Digitally Deconstructing the Constitution with Dr. Nicholas Cole

Interview published: December 23, 2021

Transcript created: July 6, 2022

Host: Jim Ambuske, Center for Digital History, Washington Library at Mount Vernon

Guest: Dr. Nicoles Cole, Pembroke College, University of Oxford

Jim Ambuske: Hello everyone I’m Jim Ambuske and this in Conversations at the Washington Library. When delegates assembled in Philadelphia in the summer of 1787 to write a new constitution, they spent months in secret writing a document they hoped would form a more perfect union. When we talk about the Constitution, we often talk about the Virginia Plan, the Connecticut Compromise, the 3/5th Clause, and other major decisions that shaped the final document. What’s harder to see are the long days the delegates spent haggling over numerous proposed amendments, precise words, phrases, and ideas that contorted the Constitution into its final form. It's a process that created many of the political institutions that we too often take for granted these days. 

On today’s show, Dr. Nicolas Cole joins me to chat about using the Quill Project to demystify the past moments that shaped our political and legal futures. Cole is a senior research fellow at Pembroke College at the University of Oxford where he is director of the Quill Project, a digital initiative that investigates the historical origins of some of the world’s foundational legal texts. And as you’ll learn today, little moments in the constitutional process can mean a lot. With this episode, we close the books on 2021. Thanks so much for joining us this past year, we appreciate the opportunity to be in your ears and we look forward to see you in 2022. Have a safe and happy holiday season everyone!

I was curious about the early part of your career when you, we'll talk about Quill Project and constitutionalism and lawmaking in a bit, but you started out your career as a class assistant and, interested in classical reception in the American Founding. Is that right?

Nicholas Cole: Yeah, that's correct. Uh, I read a degree called ancient and modern history at Oxford as an undergraduate and started with Miriam Griffin who became my, my doctoral supervisor.

And I had to write an essay on Cicero for one of her classes. And it happened to fall on the day of the American elections, I couldn't resist and I ended up writing a sort of undergraduate essay about Cicero and America. And Miriam was American herself and really liked this, had me back for a cup of tea at the end of the year and said, you know, if you wanted to do a doctorate, I think that would be a good topic to pursue.

So I wrote my doctorate on the use of the Classics by Jefferson's generation, as they were thinking through what a republican system of politics would look like in America. And that was what my early work was on

Jim Ambuske: There’s been some recent work again about the Classics and the American Founding.

And I'm thinking about Thomas Ricks’ book on, I think it's called First Principles. What do you see as the importance of classical thought to the Founding Generation? 

Nicholas Cole: Unlike most classicists who get a hold of this topic, I don’t really think that the American system of government was sort of shaped by classical institutions or particularly by classical reading.

I think it's a little bit more subtle than that. The Classics helped the Founder's generation to articulate was their unease about republics and how you would make a republic work. It helped them find a language to justify some of their institutions that didn't draw upon a sort of British monarchical tradition as well.

So a lot of their interest was to think through, you know, what was really different about America. Now that they'd separated from Britain in a way, not because their institutions were fundamentally transformed but in spite of the similarities with what they’d inherited, how could they nevertheless sort of articulate a difference that America would make for the world?

And so you still have on the dollar bill, which I really love, you know, the quotations from Virgil NOVUS ORDO SECLORUM, a new order of the ages. And that was really how the Founding generation saw themselves. When Virgil was talking about it, he was talking about the founding of Rome as a really pivotal moment in world history.

And I think reading the Classics helped Jefferson's generation to really think of their own achievement as transforming the world, or at least having the potential to transform the world.

Jim Ambuske: As part of transforming that world, they had to build new institutions to support their idea of this republic.

Institutional history, it was a fad a long time ago and it's coming back into vogue, I think. What do you see as the significance of studying institutional history? I guess, what does it mean to study the history of institutions? 

Nicholas Cole: It's interesting isn't it that if you had studied history at the turn of the 19th, 20th century, you would have really studied the history of institutions first and foremost.

And as you say, it fell out of vogue, and I think it's come back again. I think it fell out of Vogue, partly because people felt maybe the major battles had been won. Maybe the world agreed that a liberal democracy on America's kind of model was the way that everybody would want to be governed in the future.

And so there was no particular need anymore to worry too much about institutional history because in a way that battle was over. And I think it's come back into vogue now, partly because of tensions within European and American societies that people need to rethink. And partly because externally there are countries in the world that are now saying with a renewed confidence, we don't believe that a capitalist liberal democracy is in our future.

We think there are other ways to solve the world's problems and we want to champion other ideas instead of that. And so I think in a way, under both challenge internally and challenge externally, the historians have turned their attention back to the history of institutions, which I think is a good thing.

Jim Ambuske: I was thinking, as you were describing this history, and wondered if the high point was the end of the Cold War? When it seemed like there was a supremacy of democratic capitalism, liberal democracy, and then in recent decades, as we've seen people in, their own internally, and of course in the European Union and in Great Britain are no longer trusting of particular ways of doing, or institutions.

And certainly in the United States at the moment, there's a rampant distrust of almost everything. And so it seems like, that it's important to bring back that study of institutions and probe more deeply about the origins of maybe even our current moment. 

Nicholas Cole: I think that's absolutely right. I think what previous generations really understood is that the institutions that we take for granted took hundreds and hundreds of years to fight for and are very fragile.

And you need understanding in order to survive. But I think you're right that at the end of the Cold War, there was an enormous confidence that the American way of doing things, viewed in a very positive light at the time, would be the way forward for the world. And I think kind of complacency did set in.

I have to say my own interest in American history, you see I'm, I'm British, not American. My own interest in American history in a way was part of this. I came to America as an undergraduate during the summer vacation for the first time in, in 1999 and worked for a summer in Congress as an intern. And part of my motivation for doing that was that at the time, it really did seem like the American system of government was a kind of future for the whole world.

And it was a very positive moment in that way. And that was really why I became interested in American history at all.

Jim Ambuske: Did you see in that experience, the seeds of what becomes the Quill Project in your attempt to use digital technology to reconstitute the process of law making and how final texts like the constitution get negotiated?

Nicholas Cole: I was interested from that moment on, in the history of American institutions and the difference between the British system of government, which I think has a lot of merit, actually I'm not one of those people that thinks, you know, Britain's got it all wrong and America’s got it all right. 

But I was interested in this conceit that America has that you can write a constitution through a process of negotiation. I thought that was really fascinating. In the ancient world, they had had written constitutions, they had had law givers, but it would always be a single person. You would find the Solon of your community and say, look, why don't you have a really hard think and tell us what our law should be and we'll do that. 

And that was how previous societies that had had constitutional texts, had their constitutions written. It was by one very revered figure and America had done something really novel in the post Revolutionary period. They, they'd written constitutions in committee and I thought that was really fascinating.

So from the beginning, I was really interested in that. And the Quill Project really grew out of a frustration that I felt after many years of working on American political thought and institutional history. With the difficulty of understanding that process, the Quill Project is an attempt to understand the dynamics of those negotiations rather more precisely than what’s been possible before.

Jim Ambuske: I first encountered the Quill Project, probably appropriately enough when I was still at the University of Virginia Law Library and someone had just sent me the link and said, hey, you should look at this. Cause we were working on some legal history projects and somebody thought I would be interested and, of course, I was.

You just alluded to what Quill does, but I was wondering if you could speak a little more broadly about the project itself and its conceit? 

Nicholas Cole: Let, let's start with what we know about the Constitutional Convention. It's a three to four month process, and as people typically tell that story in American classrooms, the Virginians present the Virginia Plan and New Jersey presents the New Jersey Plan and they kind of split the difference and that's the constitution and everybody goes home. 

And if you give that exercise to a group of very clever high school kids, they can split the difference in about 45 minutes and call it a day. And so you have to ask the question, what was the Convention doing for the other three and a half months? You know, and looking at the records, I was struck by the nature of them. The official journal, which most historians are less familiar with than Madison's diary, doesn't record any speeches at all.

It doesn't record any of the famous quotations that we think of when we quote the Founders. It's a record simply of proposals made and decisions taken upon those proposals. It's quite dry to read. It doesn't record any of the sort of human drama, but it does record this process of going through a series of documents, iteratively line by line, making a series of small changes.

And it's also very difficult to read because it's a huge task of memory to try and think, well, if they've agreed this, this, and this. What is the text that they're actually working with? Now, what the convention had in the room was a secretary, William Jackson, whose job it was to, first of all, keep a record of what had been proposed and what had been decided, but also to be able to read out for members of the Convention at any moment, what the paragraph under discussion really was.

And that's the thing that we're missing when we come to these records. The idea behind the Quill Project is kind of simple, which is supposing we let the computer show us a reconstruction at any given moment of what the text under debate was. And supposing we tell the computer what all the proposals were on the orders in which they were taken and when votes were taken on them and then let the computer tell us well, given all of the proposals made and all of the decisions taken [00:10:00] the state of the documentation in front of the Convention, as they're making this particular choice right here is this.

 And so that's what the software platform does. 

Jim Ambuske: It seems then what you're saying is given the amount of extra material that surrounds the Constitution, besides the official Convention record, you know, Madison's Notes, you've got various letter writers writing back and forth, talking about various developments. You have delegates themselves bargaining and negotiating, that it would be very difficult for a human to reconstruct all of this absent of computer assisted technology?

Nicholas Cole: Yes. And in a way, it goes back to my classics background because after I'd done my degree in ancient and modern history, I then read a master's degree in Greek and Roman history and learned Roman law at the same time as doing that. And what's striking is if the Classics faculty had the kind of materials available to historians of the Convention, they would have produced a bookcase full of commentaries on every dot and comma in Madison's Notes, in the journals, in all of the letters that we've got, uh, relating to the Convention, that there would be a really detailed, critical commentary on all of these records.

And I started to wonder, well, why doesn't this exist for the records that tell us about the Constitutional Convention? And the answer I came to was precisely that it is really difficult to follow the flow of debate. And so part of the point of producing software to let us follow that flow of debate much more precisely and accurately is just to say, okay, if we let the computer do the really difficult bit, which is help you keep track of a constantly changing text or set of text, then we can start to write much more accurately about the decisions being taken. 

We can see speeches literally in as close to a precise reconstruction of the context in which they were made that we will ever be able to get. And so we will be able to be much more precise then, as we write about particular suggestions made at the Convention and why they were made at particular moments and not others. 

Jim Ambuske: It seems like then you can re-inject the human drama of the moment, as opposed to you reading the journal, which just, you know, this was done, this was done, this was done.

And now you actually get a better sense of probably what Madison was seeing as he's watching all of this. And as close as you possibly can to reading the reading look on people's faces in some respects when they’re objecting.

Nicholas Cole: Absolutely. And that was really our hope that, you know, having, having been as scientific as possible about how we dealt with the different records, that that would then let people, as you say, see the human drama differently again. 

And I should say, you know, in producing the digital edition that we have online of the Convention, we've used every single record available to us. So not just the journals, not just the official papers, but everything we have, we've laid out as accurately as we can. I think all of the information currently extant about the convention.

Jim Ambuske: Tell me a little bit about that process, essentially this document hunt, cause we've got published records of the Convention. We've got published records of letters from the delegates, things like that. So there's a whole corpus you're working with, but before we even start writing a piece of code, I'm imagining that you probably went out and began to see or define every last blessed piece of information or documentation you could to construct the ecosystem that the platform would then interpret.

Nicholas Cole: Well every, every generation builds on previous generations.

And of course scholars of the Convention, the Federal Convention, have been very lucky to be able to build on the work of Max Farrandd from the early 20th century who had done a lot of the collation of extant manuscripts for us. So there were a few manuscripts that came to light, but really not that many. 

So in terms of building the original software platform, we started really with what was in Farrand and then moved out from there. Which was a huge relief, because if we'd started with the sort of state of the records for some of the state conventions that we're now looking at, it would have been much, much harder to build a software platform at all. But Farrand had done a really good job of collating the records that were available.

Uh, so we started there and then the task was really, uh, before we wrote a single line of code to think, well, what would a, a model of this process look like? And I remember there was one point where I sort of had piles of papers in my office where I was trying to think through, well, if I was the secretary of the Convention and I'd been given the Virginia Plan, and the journal entries I was writing look like this, what would it be in the process from getting from their original written documents?

Through the events in the room, to the record that I then put down in the official journal and thinking through that process then helps us think through what the computer model would be and then we sort of coded up a model for debate based on that, based on the manuals of parliamentary practice from the 17th and 18th and 19th centuries, so that we had a sense of how the participants themselves thought about this kind of process of negotiation and built that model into the computer so that we could then put the data from the source material in and have it do the right thing.

Jim Ambuske: In order for the computer to read the material then, I imagine then you're probably taking what we call the full text layer or the text layer, essentially of a transcribed document or a published document to perform analysis. You probably didn't have to do a whole lot of transcription I imagine? 

Nicholas Cole: You’re right, not for that project because the Max Farrand transcriptions were good enough for our purposes.

We've subsequently gone back and done some revisions later with more recent material, but at least in terms of a kind of proof of concept, it was okay to, to use the Max Farrand material. Even though we knew there were certain readings of the manuscripts that maybe weren't quite right or that we would want to revise later, it was a huge relief that we could use that material and that it was out of copyright and that we could focus on building the software platform without worrying about all of those kinds of issues. 

For later projects that we've been working on, we haven't been so lucky and we've had to spend a lot of time doing our own manuscript work and our own transcriptions, but you need a proof of concept when you're trying to make anything like this work.

And we were very lucky that for the 1787 Convention, the records had been taken to that state where we could get on with doing the sort of difficult bit first in terms of the computer modeling. And then we would have a system that we could apply to other cases. 

Jim Ambuske: Let's talk a little bit about the process of building that proof of concept.

What does it actually entail to build a digital humanities or digital history project like the Quill Project? Can you tell us a little bit about challenges you faced, uh, and some of the decisions you had to make as you were building a computer program to interpret one of the most significant moments in American history?

Nicholas Cole: I suppose, the first thing was persuading people that it would be worth doing at all. This is sometimes one of the biggest hurdles people face, you know, what were we going to learn from doing this? And I have to say, even as we started out, I did have that fear, you know, maybe we'd put all of this effort into building a computer system and then maybe it wouldn't tell us anything very much. That kind of worry must haunt people when they start out on projects like this.

In fact, I used to ask colleagues, guess how many proposals had gone into writing the final US Constitution. You know, how many things had the convention voted on? You know, approximately how many votes or how many decisions did they think the Convention had taken? 

And people would guess about a hundred, about 200 votes. And actually when you read the records carefully in order to build this kind of model, and when you really do track every time the Convention made a decision on anything you discover that they had about a thousand different chunks of text that they took individual decisions on in order to build the final text.

So that our, sort of timeline of the events that make up the Constitution is about 2,500 events representing proposals made or decisions taken at different moments. So approximately one proposal or one chunk of text being voted on for every couple of words in the final text, which gives you a sense of the sort of level of detail that the Convention was actually working out.

So one challenge was persuading people that this would be a useful thing to do. The next challenge was working out first of all, what was the precise task that we wanted the computer to do? So it's all very well to speak in general terms about what we'd like a system to do x, but thinking through where we want each amendment to be represented in the database precisely once.

And we want the computer to then be able to produce arbitrary versions of documents that either include or exclude specific proposals for any given moment in time. And we want to be able to visualize it. We want to be able to search it in particular ways on site. And so thinking through the requirements of the project and then finding the right technology to actually achieve that.

And this for us is an iterative process as well. We've had to develop some techniques ourselves that are not used anywhere else in the world for anything, because nobody's ever been interested in modeling this kind of process and this level of detail. So we, we've taken some techniques that are used in other fields and extended them for our purposes.

We needed to think through how people were going to put data into the system and what were the skills we were going to require of people that were going to put data into the platform. I was very, very keen that they should be able to focus on being historians. I had a student in mind quite early on.

She's just finishing her doctorate, now a Grace Mallon, and she's not technical at all. So if I gave her a system and I said, she was finishing her undergraduate degree, and I said, would you like a summer job helping us to sort out the records of the Constitutional Convention, but she's not technical at all.

If I had said you need to learn all kinds of squiggly computer code to get started, I don't think she’ve wanted anything to do with the project, but I knew she was a really careful reader of documents and would be really precise, which was exactly what we needed. So we needed to create an interface that would let a historian be a historian. 

And then for readers and users, we needed to think through, well, what kind of user interface would people want to get information out of the system? And, and that continues to be a bit of research that we constantly do is, we try and meet the needs of different users for different projects.

So there are quite a lot of moving parts to a project like this, and that's before you get into the just history side of what records do we have available for particular processes and how are we going to use them and how are we going to sort out discrepancies between them and all of that kind of, sort of more traditional historical work as well.

Jim Ambuske: I liked the way you described the relationship between the historical detective work and the computer. I mean, I, I think sometimes, I mean, I've encountered this working in this field that, you know, folks get worried sometimes that the computer is going to replace them in some capacity. And that it'll take over the work of doing the interpretation.

Then it'll essentially be like The Matrix and we will just be batteries, powering the machine. But in reality, these are tools designed to assist us in our interpretation and that they can help us advance our understanding of a particular topic, like the Constitution, as opposed to delivering it, like almost Solon, like, or like Lycurgus like from on high and calling it a day.

Nicholas Cole: Yeah. And actually one of the biggest problems we face is that people are spoiled by technologies like Google. You can type into Google, where is the best place to get a cup of coffee near me? And Google will give you an answer that you're happy with. And you can't, you can't even begin to think that a system like that for legal interpretation or historical interpretation would ever be possible.

And actually modern users have a real problem. We often think of users as real technophobes, but actually users are so used to being able to type in a question and get an answer that satisfies them immediately that actually saying to them, well, look we can give you a lot of information that you can use to analyze the problem that you're working on better.

That's almost become, counter-cultural. Actually people expect a computer system to be able to give them an instant answer to the problem they have. Whereas I think the Quill Project sees its role as giving people better tools to do more precise interpretation, or more precise analysis of whatever their problem that they're working on happens to be.

And so absolutely. I see computers as a really important tool, um, in history and one that should be used more effectively by historians. And I think historians need to think through more carefully what computers can do for them and how computers can help them do things that they simply couldn't do on paper in the same way.

But I don't really like the term digital humanities. It sort of suggests that there's this sort of field where people are using computers to do clever computery things. And then historians can just get on with being historians, as they always have been. I think modern historians need to use computers, modern historians do all use computers, whether they like it or not.

And we just need to think about how we use that tool more effectively.  

Jim Ambuske: It's almost sort of reminiscent, as you were talking about the myth of the digital native of that old Arthur C. Clark quote, where a technology is indistinguishable from magic or something to that effect. And sometimes you gotta, you gotta learn what the magician's doing to actually pull off the trick.

And in that same line of thought, then in building the Quill Project, how do you go about putting text into a computer system that you can analyze? What's the process? And you talked about Grace's role in being very precise in these documents and I imagine there's a part of the work there is a human element where students such as Grace and others have to manually extract some of the texts.

Nicholas Cole: Our flagship work is the 1787 Convention, but our, but our bigger project is how does American constitutional law as a whole get written? America's obviously got 50 states now, but between them, they've all had 235 different constitutions written roughly from scratch by convention processes. So this convention process for writing a constitution is a real American institution.

And for most of those, we're not lucky enough to have carefully collated, edited records with good transcriptions that have been poured over by generations of scholars. For a lot of these processes, the official record is a very hastily edited journal that will have been published for whatever reason.

Possibly decades after the event, maybe some records stuffed somewhere in a corner in a state archive, some newspaper accounts, because most of these conventions were public, even though the federal one was private. Uh, occasionally if you're lucky correspondence, private papers, but a, but a real mixture of different types of evidence.

So our first task, if we're looking at sort of state level constitution writing, is to pull all of that material together to organize it, to Transcribe it so that we can use it. And then to go through each day of the process, most states, about the average is to take about three months overwriting a constitution.

Couple of states do it very quickly in a couple of weeks, they just borrow next door's constitution and call it a day. Some states have taken much more time on occasions, but call three months about the average. To go through each of the days where we know that either the main convention met or the committees and see what information we have about what went on on each of those days.

And what all of these processes have in common is normally a concern to keep a very careful record of what was proposed and what was agreed, at least for the purposes of the members themselves. They might not have looked after that record very carefully afterwards, but that's, that's what you needed on the day.

And so we have our researchers go through each of the days of the convention and see what we have and where they meet gaps in the record, then that can inspire them to go out and hunt in other newspapers or other places where records might exist. So actually the process of entering data into the platform becomes a driver for more archival research. 

And we ended up with a kind of virtuous circle, I think for every single project we've done, except for the Federal Convention where, where this process of collation had, had been done for us. We've started with a set of official records and people have said, there you are, that's all you need. 

And we’ve always had to go out to more material, always had to find other archives to use in order to build up a full picture of what went on. And people go through and they put the information in for us, they record the different proposals in the form the computer needs, and we can build up all of the other visualization and information that we need from that.

Jim Ambuske: So that's the historical side, the detective work we've been talking about a little bit. How does the backend work? And backend, for the uninitiated, is a sort of a shorthand term for underneath the hood of a computer system and all of the coding and algorithms designed to actually help a computer do its work.

What does that look like? And what is this thing called natural language processing? That is an important part of this process. 

Nicholas Cole: You can think of the system we've got as a series of layers. Most of our users interact with is, is our web interface. And behind that we have a system for providing the information that that web interface needs somewhere in that is the major engine of the project, which is what we call our document processing system, which is able to reconstruct on demand all of the different texts that we might be interested in for a particular project.

And below that we have a kind of database layer that stores all of the information. And you're right, there's a whole series of different technologies that go into that. So one technology that's really important for us is built on the kind of technologies that will let people, um, maybe they've played with recent versions of Microsoft Word in a corporate setting.

That'll let multiple users edit a document, or maybe they've played with Google Docs, you know, it's that kind of technology, but extended for our purpose. And then some of the technology that's built into Quill is to actually do some analysis of the use of language. Part of that is, is as you say, natural language processing.

So one of the things, for example, that natural language processing can do for you is take documents and divide up the parts of speech. And you might want to do this if you're, for example, wanting to generate some keywords. We know in education use, one of the things that teachers often want for their students is a kind of word cloud of the most important words at the Convention.

So how do we generate that? Well, maybe one way to give them that kind of visualization is to say, well, what are all the nouns that people talk about? In English, a lot of meaning is in nouns. So we're not interested in, in other parts of speech. Let's just show people the relative preponderance of particular nouns.

And so we can use natural language processing to pull out that question and other technologies as well. We might want to do some statistical work on voting patterns. There's a module that will draw you a principal component analysis of voting by individuals or delegations of the convention. So, Quill Project tries to build as many different types of tools into its platform as possible to serve particular purposes.

Jim Ambuske: And part of the process of building this, as you've mentioned earlier, is resting on student labor and even working very closely with Utah Valley University, their Center for Constitutional Studies. Tell us a little bit about that collaboration and its importance for the Quill Project. 

Nicholas Cole: That’s been an amazing collaboration.

It really began almost as soon as the project began. Utah Valley University is an open enrollment university in Utah. So it couldn't be more different from Oxford University, which is where I work. You know, Oxford University is a, is a highly selective university. It's 800 years old, uh, Utah Valley University isn’t.

It's an open enrollment university, I think it got university status maybe 15 or 20 years ago and builds on an older, older college there. But the two institutions couldn't be more different. But I happened to know the former college roommate of the university president. I got a phone call from the university when the Quill Project had just started. Would I come out and meet their students at the center for constitutional studies and see if there was any work we could do together?

And I'll admit I was really intrigued. The students seemed really, really keen to help on a project to do with constitutional history and the university had an ethos of what they call engaged learning. So finding undergraduates projects to work on, and a lot of the students at the university or first-generation or non-traditional students in various ways.

So I was, I was really keen that we would find a project that they could help with. And they helped us initially by doing a lot of bug testing, a lot of menial work, a lot of the more traditional tasks that you might associate with a digital humanities project, quite a lot of keywording, manual keywording of different things in the platform.

But after a while they said, well, we'd really like a project that we can call our own. So I said, well, okay. We've got the basic system running, why don't you look at the process of writing Utah's state constitution? Um, go to the state archives, see what records are there. See if the model that we've got will work for that late 19th century process.

We're looking at the 1895 convention. And I didn't really know what they'd find. I'd never looked at the state archives of a 19th century state. I didn't really know what the records would be like. And the students started work on that project and it proved to be absolutely fascinating. And what I thought might be three, four month project turned into a two year project to really sort out how this state had written its constitution in 1895.

And really alerted me to the fact that actually there was an awful lot of work to do at state level in terms of sorting out the history of American constitutional law. And moved on from that project to look at the archives in neighboring states, just to see if Utah was unusual in some way and rapidly discovered it really wasn't.

So Arizona State University, inspired by that project, we got money from the state legislature to work on the history of Arizona's constitution. UVU won a grant from the National Endowment for the Humanities to expand their work on constitution writing in the Western United States. And then students from that center have been helping us at the moment with the project that I'm perhaps most excited about, which is an analysis.

Uh, publication and analysis of the records related to the 13th, 14th, and 15th amendments. So the Civil War and the end of slavery in the United States. And we'll have that project coming out next year, but it's been really, really brilliant to work with the students from that university. And actually now I can't really imagine the project without the help of students, not just at that institution, but at several institutions in the United States, they've really become a core part of the project, which we'd never planned. 

Jim Ambuske: That's a nice, happy accident. It sounds like it's pretty fantastic. What are some of the benefits the students have gotten out of it? 

Nicholas Cole: I think they will come away with a well first and foremost, a profound understanding of the nature of historical archives and the political process.

And a much richer understanding of state and federal constitutional law. Many of them have gone on to law school or political science graduate work or history graduate work, and say that they wouldn't have thought of doing those things except for working on the project. So we're really proud of that.

Jim Ambuske: In all this work, what are some of the significant findings you say, have come as a result of probing these texts and reconstructing the process of their creation?

Nicholas Cole: I have a much better appreciation of the nature of these kinds of formal negotiations. I think that’s a process we need to study in its own right.

I think we'd have tended to gloss over or tended to talk about in, in heroic terms, as we look at the sort of particular battles that Madison at the convention or other figures and I think the Quill Project helps us focus on the fact that actually there are dozens and dozens of people involved in each of these processes.

And we can think about the dynamics of these relationships and how they shift across time. So one of the most important things I think is to refocus our attention on the nature of compromise in these kinds of processes and how, how, compromise works. Um, for the 1787 Convention, there are some [00:34:00] surprising findings in some ways.

The records are not as a historian would wish, you know, we know that the, the journal, particularly towards the end, gets quite untidy. As, as Jackson finds it hard to keep track of what's going on. Mary Bilder's written about the nature of Madison's Notes recently, and issues with the manuscript there.

And so on. So one of the questions we had when we started the project was if you lay out all of the information we have about what the convention debated and all of the decisions we think they took, what is the text that pops out? Is it actually the text of the Constitution? I, we didn't really know when we started the project and it turns out it is.

Not that you can get there from any one source, but with the sources we have taken together systematically, we know when every single clause in the final Constitution was considered, at least by the plenary session. And when the decision was taken to adopt it. And that’s really interesting. We also know that the sources don't disagree about what the text was.

Um, I spent a couple of months once, anticipating that the sources we had would disagree about the wording of particular clauses and thinking about how we would visualize that and building a system where the computer would be able to say, well, if you're following Madison, this clause reads like this, but if you're following Jackson, this clause reads like this.

And that was very, very complicated to get right. It was anticipating a problem though, that we never hit because although the, although the sources disagree sometimes about the precise sequence of events and you need to applying traditional historical techniques, really just reconcile a difference in ordering where Jackson for neatness is sake might've put the outcome of a vote next to the proposal that it went with rather than in the strict temporal sequence, that kind of thing. 

Um, there's a little bit of disagreement about ordering. There's a little bit of disagreement about, ordering you know, abbreviations spelling, but there's no substantive disagreement about wording at  all, which is really disappointing because that bit of computer code is, is really clever and, uh, it would have been great to deploy it, but that's, uh, that's interesting.

And then at the state level, as I say, really, the more we look at these records, the more we discover there is more stuff to say more, more things to find. There are many voices now calling for people to pay more attention to state level constitution writing. And I think they're absolutely right that the records of state level constitutions now need the kind of attention that the Federal Conventions had for a century and the state level records just haven't.

Jim Ambuske: Well I bet that a computer could, um, come in handy someday. I suspect not all is lost. How are folks using the Quill Project right now? Are they using the classroom or they, are lawmakers using it to analyze past decisions?

Nicholas Cole: It first and foremost, is a history platform. I'm a historian first and foremost. Um, I'm just interested in how these processes work historically and everything else is secondary, but we know that lawyers are using it.

We know that the Utah records have been cited before the Utah courts. And we expect that kind of thing to continue. And we know that the Quill Project is used in undergraduate courses. And we know that teachers are looking at some of the resources to see if they can build better lesson plans in school curriculums, and really encourage students to really wrestle with primary sources rather than textbooks.

And we're really, really excited to facilitate that. I'm not an expert on American high school education and I never will be. And you know, again, the Quill Project’s a very collaborative endeavor and our job is to give teaching professionals the best resources that we can. We try and serve the needs of those different communities, you know, historical researchers, legal professionals, and education professionals.

Jim Ambuske: Where do you see the project going from here? 

Nicholas Cole: So we have a lot of work to do on state level constitution writing as I've hinted, uh, that will keep us busy for a long time, very interested as well in early federal legislation. And that's a very important issue. Uh, it's, it's very, very hard to talk about things like the Bill of Rights completely separately from early federal legislation.

And I think there's work to be done there. And then we have a few international projects. So we've, we've done a little bit of work on the records of the French Revolutionary Assemblies with students from French universities, France is having a revolution at exactly the same time that America is having a revolution and writing its own version of declarations of rights of man and so on.

And it's interesting to see the different way in which parliamentary style process was used in America to France. And then perhaps for some more modern states as well. This process invented in America of having a Convention that considers documents iteratively across a quite a long period of time in, in quite a lot of detail, turns out to be one of the most popular ways of writing a constitution in the modern world/

So India in 1946, when they become independent from Britain, has to write a constitution for the world's largest democracy. And the process they adopt is strikingly similar to convention style processes that are familiar from America. So we're very interested to look at the way in which this American way of writing a constitution is used internationally as well.

Jim Ambuske: It sounds like we've got plenty of work ahead then.

Nicholas Cole: I think so.

Jim Ambuske: More Conversations after the break.

Brenda Parker [Intertwined Ad]: Over the course of their lives George and Martha Washington enslaved more than 577 men, women, and children at Mount Vernon. They labored for the Washington's comfort and wealth while building families of their own, resisting their enslavement, and self emancipating when the moment was right.

This is Intertwined: The Enslaved Community at George Washington’s Mount Vernon. Look for it wherever you get your favorite podcast, learn more at georgewashingtonpodcast.com.

Jim Ambuske: Nicholas, what book are you reading right now?

Nicholas Cole: I'm reading two books right now, but they are related. The first thing I'm reading right now is a book called The American Political System by DW Brogan and its Brogan, who was at the LSC and then at the University of Cambridge.

And he was writing in the 1930s, originally for a British audience that really didn't understand America at all. And this is one of those classic books trying to explain America to Britain. And it's really interesting what his perception of America was and what his concerns were. He spent a year at Harvard as a graduate student, uh, and come back to Britain.

And this is his major work that made him famous. He thought the biggest change to have befallen America was actually the election of FDR. That that was the biggest constitutional change, which just tells you how perspective shift, you know, the, the, uh, the, the, the end of the end of slavery he glosses over and spends a lot of time talking about, about FDR.

 And then, um, Gary Gerstle. Now he, Gary Gerstle is American, uh, but he's, he's now a professor of history, of American history at the University of Cambridge and wrote a book, Liberty and Coercion: The Paradox of American Government. So a very similar theme, sort of modern, modern take on the American system and the, the role of the American states and the federal government and how they fit together.

Jim Ambuske: Who is the author you most admire?

Nicholas Cole: Well, it's difficult. The historians who inspired me were, were Miriam Griffin, my, my doctoral supervisor we've talked about. She was a Roman historian, wrote about Roman emperors and Roman philosophy. Uh, Christopher Pelling, my first tutor, uh, writes about Greek historiography.

I love their writing. I like the prose of historians like E.P. Thompson, or Edmund Morgan, really rich analytical prose, but also just incredible to read. And then intellectually, somebody like John Pocock, History of Political Thought and Methodology, owes an awful lot to a historian like John Pocock.

Jim Ambuske: Over the course of your career, what's the most exciting document you've discovered?

Nicholas Cole: I think in, in keeping with the sort of Quill Project’s collaborative, uh, approach, it's not a document I've discovered at all. It's not even a very interesting document, but it does tell you something about what doing collaborative history with undergraduates as colleagues can really do, because there was a time that I made a Skype call to the students at UVU.

And they said, well, we haven't made much progress this week. And I thought, well, that's fine. And they said, well, the problem is we were looking for a particular subcommittee report at the Utah State Convention, uh, on, I forgotten what it was, It was, it was, it was land rights or something, but it, you know, it doesn't really matter, but the point was, it was missing from the state archives and they couldn't really make any progress with understanding the work of that committee unless they found it.

And it had gone from the state archives, they looked in the newspaper that they normally used when, cause they had newspaper accounts for most of the convention, but the, the single surviving copy of the newspaper that they normally consult had a big hole in it because somebody had cut out a, a coupon from the other side of the page. 

And so there was a big hole in the newspaper where this committee report should have been. And, and because of, um, because of photography in the 1980s and 90s, you know, most of these state level newspapers have been destroyed. There's now normally one copy surviving, and this was the only surviving copy and it was, and so, so it was a huge gap in that.

So they, they actually ended up seeing if any other newspapers of the period had printed this committee report and it took them over a week to go through different sources and they eventually found a copy of this report. And I'm not sure it said anything interesting at all, but the point was it could be recovered.

And I think that story will just always live with me as just showing me how dedicated students can be if you set them a task and say, you know, really find out what happened, they will exhaust every possible avenue. And I love that story

Jim Ambuske: Golly, that's impressive. That's persistence right there. 

Nicholas Cole: Absolutely. 

Jim Ambuske: When you ascend to the great archive or library in the sky, how do you hope people most remember your work or take away from your work?

Nicholas Cole: Oh, I don't know, that's a difficult question. 

Um, I hope they are, I hope they'll remember that we ran a hugely collaborative project that involved dozens and dozens of people. I hope that we'll leave something that's foundational for other people, both in terms of substance and in terms of methodology that they'll be able to build on.

And I hope we'll have been part of refocusing attention on legal and constitutional history and institutional history because as we said at the start of this conversation, I think there was a period when we became very complacent about the future of our institutions. I don't feel complacent anymore about the future of them and I, and I hope other people won't and that we’ll show them that there's lots more to be found out.

Jim Ambuske: Hope is a good place to end, Nicholas. Thank you very much. 

Nicholas Cole: Thank you.

Jim Ambuske: Thanks for joining us today on Conversations, a production of the Center for Digital History at the Washington Library. I'm Jim Ambuske, your host and producer. Jeanette Patrick offered editorial assistance with additional support provided by Mount Vernon's Media and Communications Department. Our music is “Witches Brew” by CK Martin. Be sure to subscribe to Conversations on Apple Podcast, Stitcher, or wherever you get your favorite programs.

If you like the show, please rate and review it on your favorite podcast app. Find this and other episodes by heading over to our website georgewashingtonpodcast.com. Thanks, and we'll see you next time.