Dear All,

I'd like to welcome everyone to the discussion thread for Panel 2 and to thank Kyle, Matt and Katie for two fascinating talks about a number of very interesting projects.  As panel chair, I have the privilege of asking the first questions, and I'd like to begin by inviting each presenter to answer one of the questions raised by the other paper.

So Matt & Katie - how would you respond to Kyle's question concerning the ways in which we might open up our projects for "serendipitous evolution" of the project beyond its original vision?  What sorts of serendipitous evolutionary developments do you anticipate/hope will emerge out of your project?

For Kyle - I was wondering how might Matt & Katie's data be brought into dialogue with your projects?  More specifically, are there aspects of your projects which might help Matt and Katie assess the distinctive Scottishness (or otherwise) of their data? 

A final question for Matt & Katie and Kyle - which emerges out of Mark Towsey's Keynote; Mark identified the very different interpretive yields generated by the quantitative and qualitative dimensions of the data and I wondered if you had any thoughts about the interplay between the quantitative and qualitative dimensions of your own project?

Thanks very much, Kyle, for your interesting and stimulating talk! I particularly enjoyed hearing the ways in which the project had developed in its different iterations, and the particular challenges that each new set of evidence posed, and the solutions that were chosen. Very interesting to think about these projects in terms of generations, and the ways in which such generational shifts can be both helpful and potentially confusing to users.

That question of keeping projects going after the funded period is of course a good one (and a vexed one!). Matt and I have been talking about how this might happen over the past few months. We are hoping, of course, to access some forms of follow-on funding, and we have ideas as to how our project might develop in two different ways - either geographically (widening the scope beyond Scotland) or chronologically (thinking about library records within Scotland but on into the C19th). And we are extremely keen to try to use our data in conjunction with those of other projects to do some comparative analysis as well.

The demands of particular funding bodies may eventually determine which way we go, though we'd be very open to hearing advice about which direction people at the workshop think might be more useful/interesting.

And of course it's much more likely that we won't get any more money - in which case we'll need to be thinking about other ways to keep the project going. We've been debating the pros and cons of crowdsourcing, as a team, and we've been keeping a close eye on developments in MS transcription software, such as Transkribus, to see whether we might eventually be able to make some significant savings of time by using such software. If so, then we could probably keep the project going ourselves, by simply continuing to transcribe records that aren't going into the initial database. Editorial time will, of course, be the issue either way. We are also thinking about how PhD students across a variety of disciplines might be able to use our data, and planning applications for funded studentships that will take the data analysis forwards.

In terms of more serendipitous evolutions - we're very much hoping that once the data is made available to the public, we will be able to tap into the expertise of members of the public who have a great deal more local knowledge to fill in some of the gaps in our data about borrowers, in particular. Local History Associations, for example, are made up of people who probably know a great deal more than we do about many of the people in our database. And our experience so far suggests that it's entirely possible that there are indeed many MORE C18th records out there across Scotland that we don't know about yet - if that is the case, then I am sure further evolutions relating to those may come about...

 I'll come back to Mike's second question (about quantitative and qualitative yields), for which many thanks - it's a fascinating thing to think through - in a separate post.

Kyle, lovely talk on the challenges of long-running archival programs (by "program" I mean a succession of projects, each built upon the last and centered on the same, perhaps expanding) set of code and data.

Often (usually?) when I hear people talking about sustainability, they're talking about continued hosting, and the minimum amount of maintenance to mitigate the inevitable code rot which sets in over time. But I almost never hear them talk about the people on the project, the knowledge and interest they embody, and the certainty that they will eventually move on. Lately, I've started thinking about this in the context of some the long-term programs I've been involved in, especially as my eventual retirement, still a few years off, has nevertheless emerged above the horizon.

Is this a point you've thought about? If so, I'd be interested in hearing your thoughts . . .

@steve_pentecost

Ditto on Steve's point and question.  There's also a sense of being trapped by the sense of obligation to maintain these resuorces, taking time away from other work. 

Thanks to @katiehalsey and @Steve_pentecost for setting the ball rolling.  

@katiehalsey raises the question of how/whether PhD students might be able to bid for funded studentships that will take a given project forward & I wondered if anybody at this conference has any examples of this that they'd be willing to share?

In the UK context, links with Local History Associations surely makes sense in terms of the 'impact agenda'?  Again, is there anybody here with relevant experience to share?

@steve_pentecost raises an important point about the expertise and knowledge which frequently gets lost (or at best under-utilised) as we move away from projects in time - the human equivalent of code rot perhaps?  Does anyone have any suggestions for ways of mitigating this problem?   

Dear All,

Starting a new topic here, based on something which sprang to mind during Kyle's talk and which resurfaced during Matt & Katie's talk - which is whether there's anything to be learnt from the construction of Gothic cathedrals here?  I'm thinking (in a quasi-Ruskinian way) that Cathedrals provide an analogy for thinking through the problems raised by multi-generational projects.  I'd also be willing to wager that serendipitous evolutionary developments also played a role in those projects too.  Should we opening discussion with medievalists about such things?

These great questions and observations about sustainability also connect with some themes in Panel 6 and the RED project's successful evolution and management of leadership succession over a 25-year period. 

Thinking of @MikeSanders's point about Gothic cathedrals, there's also a way to see universities themselves as adapted medieval institutions that can sustain certain kinds of communities of knowledge across generations. 

Apologies for the delay in responding - it's been a very busy week!

On the sustainability issue, I think it's quite important for projects to think of themselves as producing two different things: good, reliable data and an approachable interface for exploring this.  With Books and Borrowing, we're hoping that the system we build will be available for a long time, but we're also keen that the project produce good transcriptions and interpretations that will be deposited in key repositories and with our partners so even if our system ages out, the data can be taken up by other users and projects to be improved and carried forward.  There's a certain satisfaction in getting the data from manuscript to digital form, but it's also painstaking work - once we've done it, we don't want someone else to have to do it again!

@mikesanders Thanks for great questions. I think the comparative angle on student reading could be really interesting, both for what it might say about what is distinctive about a dissenting education as well as what is distinctively Scottish. Some Dissenters felt more comfortable at Scottish (and also Dutch) universities than the English ones, so I think there are ways they might actually be more similarities than differences. I'd love to see an overlap study to get a sense of 19th century student reading in general, which could, of course, extend to the Continent and the United States as well. My other project from the last decade is on Jesuit libraries, so we might, in time, be able to tell a much larger story about higher education over the long nineteenth century ...

More of the early analysis on Dissenting Academies borrowing (much of it still unpublished) tried to mix both, but a decade ago the open source visualization tools were not as easy to use as they are today. The qualitative approach in some ways was easier, but the issue of the representative reader always looms. What I appreciate about the Dissenting Academies data is that the vast majority of records are for students who were enrolled for several years. So you have this large amount of information about a group of young men experiencing a stable curriculum for a relatively short period of time. So with the right application, you can isolate what you think / know is assigned reading for class and discern what is personal reading. There will be more deviation in the latter, but then you can also see the kid who came into school wanting to be a local minister catching the mission bug! And there is a massive amount of periodical reading happening, which helps expand us beyond library histories' focus on the book.

Your question is also reminding me of a project that a former colleague and I had pitched a fellowship project (alas, unfunded!) to visualize individual borrowing histories as pathways through the library, utilizing the shelfmarks to reconstruct the physical locations of books within the collection, so that you trace the movements of readers physically within the collection. I saw an installation this past winter by the artist Spencer Finch which included his carefully observed movement of bumblebees among flowers ( https://www.berggruen.com/exhibitions/spencer-finch?view=slider#3) and thought about how I would love to do that with different readers movements among books. (I'm a little obsessed with serendipity.)

In sum, there is so much to do with this data BUT so little time when 95% of our time on grants is in constructing it and we are then need to find other projects / positions to pay the mortgage!

@katiehalsey I feel like the type of position we are in when we undertake our projects plays into what follow-on looks like. I so, so loved hearing in Matt and your presentation about the broad number of students you had been able to involve in the project. As everyone in the workshop knows, library history is remarkably generative for students. The Jesuit Libraries Provenance Project ( https://jesuitlibrariesprovenanceproject.com/) which I started when I was at Loyola had over 40 different interns and students working on it over a six-year period. I always said I would stop the project when students lost interest. Turns out I took another job and started a new project, although now I'm writing a monograph pulling all of that data together. The commitment you and Matt are making to bringing in partners is very much to be lauded!

Transkribus is intriguing, but if you need 100 clean pages of transcription of a hand before you can really train the algorithm to read it, you're likely going to run out of the type of library records we have to work with. It feels like a big investment that pays off big with voluminous writers (like Lyell) but is harder to sustain with smaller collections. That said the proprietary HTR seems to work pretty well on smaller batches, but it's proprietary, which is a barrier ...

I'm probably rambling at this point, but I'm intrigued by @MikeSander and @Douglas Knox's thoughts about the Cathedral as model. I don't know much about medieval cathedrals, but I do know way too much about 19th C urban evangelical meetinghouses. The congregations that inhabited the latter in New York City basically tore them down and rebuilt every generation, never valuing them for their historical associations, but instead seeing them as manifestations of the vitality of the current inhabitants. (It's only around the time of the American Civil War that shift happens as evangelical optimism and wealth wanes. It's also around this time they finally accept the Gothic as not being too Catholic.) I feel in some ways our projects are like that - interfaces look dated, we feel that if we refresh the styling we'll attract new audiences, and we want to signal the project isn't going to be mothballed. With this rebuilding comes new functionality, but there also comes losses (as I think @JulieanneLamond was saying about something she preferred more in the first Australian Common Reader than the new version). But what would it mean for these projects to be valued for their contributions and then added to by new individuals with their own motivations? Is mission creep a good thing? Can / should a project that was about borrowing records become a project about marginalia because that is where interest (either by the creator or the field) headed? I've probably strayed from cathedrals at this point, so I'll stop.

@steve_pentecost I would love us to talk more about sustainability in terms of people. It intersects with the DH projects we do in so many ways. There are projects who become so identified with / dependent on a creator that it is hard to imagine their existence without that creator. Here, I'm thinking about Jerome McGann's comments a few years back about the future of the Rossetti Archive. Dissenting Academies survives because of the fierce advocacy of Isabel Rivers, the original grant PI, but she is no longer taking students. Then there are the number of postdocs employed on these digital projects who walk away from (or not invited to stay on) after a project ends. Of the four humanities postdocs I worked with on the larger Dissenting Academies project (which included a Database & Encyclopedia), I'm the only one still really connected to the project. While postdocs, in the best situations, get excellent training and go on to do their own great work, the loss of their technical and content knowledge is great for a project, and I'm not sure we always fully appreciate that until it's too late.  (This is especially the case when we write grants which only provide time for data creation and interface development, but not meaningful scholarship out of that data.)

Another part of human-centered sustainability is documentation and reflective practice. I've not always made time to record the ways in which my thinking on a project has developed over time, even when those changes have been profound. Teaching in a leading American public history program for eight years has convinced me it's more than just journalling as I go, but developing more formally written pieces on what doing work at the intersection of humanities and commuting has done for my own understanding of what it means to be a scholar (and a human). I'm grateful to Jim and Frank for the opportunity of this conference to stop and take stock of a project, forcing myself to articulate my thoughts about a significant project in my own development and how it - as well as myself - have changed over time. For example, working in an independent research library now, issues of critical cataloguing and privacy are very much at the forefront of our conversations. What would it mean to turn those concerns on a project created a decade ago when I was not thinking about those issues? I'll have neither the time nor the resources to make those changes with Dissenting Academies, but I can make sure I do the work of participating in workshops like this, dialoguing with other longterm laborers in the field, and continuing to reflect on the work I've created and the implications of it.  

@msangster on that question of sustainability, a scholar at a Victorian Periodicals Review event earlier this week recommended creating a pdf of digitised materials as a way hedging against the loss of the website.  I'm not the most technically-minded person, so this may be a very stupid question, would this be feasible for your project with @katiehalsey ?  

@kyleroberts6 your comments about wanting to find a way of visualising borrowing records as pathways through libraries reminds me of something we discussed as a possible aspect of the Piston, Pen & Press project I'm involved with.  We wanted to create a virtual Mutual Improvement Society reading-room with library through which your avatar could roam and select books from the shelves.  Alas this would have busted the budget - but one of our digital advisers said that this was the kind of project that might be suitable for computing science students (as the movement through space element would be good re developing computer games).  All of which is a roundabout way of asking, if anybody has any experience of working with computer science students that they might like to share?

@kyleroberts6 and others, I'd like to hear a little more about your work with students.  Was this restricted to interns or did you work with undergraduate students in a way which made a primary research project a key part of the course?  This is a very self-interested question, because I've been thinking about the possibilities of such a course - but am very uncertain as to what might reasonably be expected of students by way of technical competence (to say nothing of my own limitations in this area).