OK, Fred, before we start the show, I need to give you some context. We did episode 605 on June 7, 2024. No broadcast problems, but the station's encoder was, we'll just say, on the fritz.
So there was no podcast version.
So there was no podcast. We're taking a Mulligan on this particular episode.
Well, this is going to be part one of a series. And we've already done part two.
Just roll the cold open, man. Let's get going.
As an archivist, there is one question I dread. Not, what is an archivist? I love my job and I can talk about it at length. It is, after all, looking after the written heritage of the world. Not, it must be lovely working with all those old documents. It is. The question I dread is, aren't you going to digitize it all? The answer is simple. No.
[TELEVISION, "FRICTION"]
You are listening to WREK Atlanta, and this is Lost in the Stacks, the research library rock and roll radio show. I'm Charlie Bennett in the virtual studio with everyone-- Marlee Givens, Fred Rascoe, and Alex McGee. Each week on Lost in the Stacks, we pick a theme and then use it to create a mix of music and library talk. Whichever you tune in for, we hope you dig it.
Our show today is part of our new series, the stories that archives tell, and is called "You Can't Digitize Everything," the Lyman Hall edition, which I hope let's you know, out the gate, I'm looking to break some hearts and set some stuff straight.
Alex, one might think you have an agenda for this episode.
Yeah, I kind of do. In archives, we often hear from folks that we should digitize everything and put it online. And that would be the best. But the reality is so much more complex than that.
And so you have brought some backup.
That's right. To talk more about the complexity of digitizing everything and a recent mass digitization project we are doing here at Georgia Tech. I've invited my colleague and resident digitization expert, Katie Gentilello.
And our songs today are about easy solutions, solving plans, and telling stories. As Alex said, a lot of people think the solution to archival and access problems is to just digitize everything and pop it on the world wide web. Easy, right? Well, that's an easy to understand idea, but it's very wrong. It's harder to explain why we can't digitize everything. So here's "Hard to Explain" by Coriky, right here on Lost in the Stacks.
[CORIKY, "HARD TO EXPLAIN"]
That was "Hard to Explain" by Coriky. Our show today is "You Can't Digitize Everything," the Lyman Hall edition. And we are joined in the studio by current and future guest Katie Gentilello, Digital Project Coordinator at the Georgia Tech Library. KATIE GENTILELLO: Thanks for having me. I guess we're going to start with why digitizing everything just simply won't work. We know that it is time-consuming. There's a cost to it, the labor costs, the environmental cost.
But let's talk about all of that, Katie.
So I've been doing this for, I think, 28 years at last count. So I've seen a lot of different collections over the years. Lots of different materials have different levels of usefulness, I think, for researchers, for just the community in general. But there's a lot of it. And there's a lot of different types of materials. I just recently had a conversation with a friend of mine who said, well, isn't it just easy to just digitize that book or digitize that rare material?
And I explained to him just how long it took to prepare the materials, the selection, the handling. The digitization is the easy part. Then to make those images useful and adding all the value added-- metadata, description, and then the systems that they'll go into. It's a very long process. And when you multiply that times thousands upon millions of pages of documents, it really does eat up a lot of time and resources. So yeah, it's a pie in the sky idea. Wish it was a lot easier to do.
But unfortunately, it's a challenge.
And Katie, I remember the beginnings of digitizing everything many years ago, watching you take very careful digital photos of maps. So I think you might even be minimizing the digitization requirements. Because you're good at it. You do it now. But taking a picture or making a scan of page after page or objects, even that's not very simple. People seem to think that it's like, run it through some machine. But the machine is you.
KATIE GENTILELLO: Yeah, I mean, there's a lot of different digitization equipment out there. I think most people are familiar with just flatbed scanners, which is take your single sheet, flop it down, take a scan, and then you've got it. But that doesn't work for archival materials.
You have books, you have bound volumes, you have oversized documents that require special handling and special technology to be able to digitize those in a way where they're not going to damage the original, bending pages, or cracking book spines, or even down to the lights, they can't be putting out heat. So there is a technology specific to the materials that you're using. And so there's a learning curve of learning how to use it and then and then using it correctly with the materials.
Yeah, so it's yet another challenge to point that you pointed out. CHARLIE BENNETT: And you then said you have to add stuff to it. And that's the metadata, of course. And I feel like this is where we turn it on you, Alex, and start asking you questions. Do you work with digitized papers that come to you and then have to come up with the metadata, the filling in all the forms, all the stuff?
So I don't do that part usually. I do the ArchivesSpace description, which kind of influences the metadata, or the metadata can influence the ArchivesSpace description. I mean, same idea, thinking about how people find these things when they are digitized online, what's useful. That is an additional layer of description beyond just the digital files themselves. So there's the scanning and the metadata for the scans.
And then there's the description piece in our repository, our finding aid system.
Do you have a sense, Katie or Alex, how much time or money or both that one piece of paper in the archive collection scanned and then made accessible from zero to fully accessible, what kind of effort, money, resources are we talking about?
This is a thing that everyone is trying to quantify. There's lots of calculators and things online. People have written papers about this. And it's so dependent on what your setup is and the level of detail you're doing. I mean, Katie, I don't know that we've done it here. But I mean, we have student workers. There's the cost of literally the storage space that we keep for our files. If you're talking about digital preservation on top of just the access copy, that's another layer of cost.
It's not cheap.
I mean, I can address based on the Lyman Hall letters about how much time it takes to do 50 pages. Scanning, I believe, we were doing it over several hours, so we would probably get maybe 200, 250 pages done in that period of time. Then you take those raw scans and I have to quality control check them. Then I have to run them through a filter to make them readable because they faded. The text is faded so much that you can't just look at the document and be able to read it.
I have to run it through a red filter to make it contrasty. So there's that preprocessing, getting it ready for transcriptions. Then we're hand-transcribing them because we cannot apply OCR, which means [AUDIO OUT] which is optical character recognition software to it. And then to transcribe, just to actually read and transcribe 50 pages probably takes about 10 hours.
And then after that, I follow back up and fill in any gaps that were missed and formatting issues, save it, create derivative files, and then from there those files go into a queue to be added to the repository. So you can just see that for 50 pages. And we're dealing with 10,000 pages with this Lyman Hall collection. So you can do the math.
No, thank you. CHARLIE BENNETT: I can't actually. Yeah.
[LAUGHTER]
I sometimes want to say when someone does that quick kind of, well, we just digitize everything and then you'll be able to get to everything on the internet or something. I want to ask them, when's the last time you went through and captioned every one of your digital photos?
This is Lost in the Stacks. We'll be back to talk more about the difficulties of digitizing everything or not everything after a music set. File this set under BD71.S37.
[BEST COAST, "HOW THEY WANT ME TO BE"]
[VOCALIZING]
[BE YOUR OWN PET, "BIG TROUBLE"] And when I tell you I want you to believe That was "Big Trouble" by Be Your Own Pet. And before that, "How They Want Me to Be" by Best Coast, songs about recognizing the flaws and easy solutions.
[MUSIC PLAYING]
This is Lost in the Stacks. And today's show is called "You Can't Digitize Everything," the Lyman Hall edition. We have Katie Gentilello, Digital Projects Coordinator at the Georgia Tech Library in the studio with us today. We've talked about how time-consuming, how expensive of all the processes. Let's talk about how we decide what we digitize in the archives. Katie, how do we make our decisions about prioritizing what gets digitized for a large scale project like the Lyman Hall one?
Yeah, that's a good question. There's a lot of considerations that go into how do we decide what is worthy, I guess, is maybe not a fair assessment, but because we have this limitation, we have to decide on a rating scale. We have a prioritization matrix that we've come up with as a department that looks at different factors, such as, is this material unique? Is it something that's already been digitized and is available online with another group?
How many requests do we get, research requests do we get for this type of material? What is its historic value? Do we rate it pretty high in its uniqueness and the impact that it has on our researchers? Or is it low? We also look at how diverse is it in terms of what are we offering online, what story is it telling. And then the usual ones like copyright-- is it in copyright? Is it not? Available metadata.
We keep talking about metadata, but when you're handed a collection that has very little processing done to it and you're kind of doing that in tandem while you're digitizing, it makes for a very, very-- you're adding more time and more labor. So all of those factors go into a rating system that we apply. With the Lyman Hall situation, we chose this collection outside of that prioritization matrix because it was identified as something that was at risk.
And that's also another one of our parameters. The risk of this particular collection being lost to the ages was so high that all of the other factors became kind of a moot point. And when I say at risk, these letters are over 120 years old and were typed on onionskin paper that is very brittle and it was falling apart. And so you're dealing with faded text and you're dealing with, literally is crumbling into shards.
So we were able to digitize that and then be able to augment its quality so that we could get the content. So that was a bit of a unique digitization prioritization situation. That's a lot of words. But it was necessary in this particular situation. And I'm glad. I'm glad that we did it because it's been very enlightening to uncover so much rich Georgia Tech history that has been inaccessible for quite some time.
I was just curious if the matrix also has an option for farming it out, like hiring a contract service, or if that's just something that every individual library decides if they want to do that in house or not.
Yeah, that's a good question. It's not part of the prioritization matrix officially. But it is something that we will certainly look at on a collection by collection basis. Just as an example, we have the living history collection that is a collection of DVDs, DV tapes, VHS. And it was actually proposed as like, is this a collection that we should outsource or is this something that we can do in house?
And having sent collections for outsourcing, I mean, the biggest, I think, elephant in the room with that is the risk. Your collections are leaving your custody. So then you have to make sure shipping and just the act of getting them to some other unit outside of your organization is concerning. I've also had successful outsourcing projects. So yeah, it's just a case by case basis.
Katie, let's get the listeners up to speed on this collection because we all know Lyman Hall, both as a building and a person. But what is this collection of letters and why is it important?
OK, So just as a quick recap, so Lyman Hall was the second president of Georgia Tech. And he was instrumental in fundraising for Georgia Tech, especially getting the French textile school up off the ground. He was also responsible for several dormitories being built. He was a military man, so he had very strong ideas about discipline on campus. His letters started in 1898. So he was about, I think, on his third year of his presidency. They run until his death in 1905.
They're letters from outgoing correspondence that he was able to document throughout his presidency. And they run the gamut from just, hey, here's some information about Georgia Tech to asking Andrew Carnegie for money for a library, and asking the city of Atlanta to pave North Avenue, and writing letters to Mr. Peters about the land where Peter's parking lot is, that he was trying to get some of that land purchased for Georgia Tech. So there's a lot of rich history in there.
And every now and then, I'll uncover something I get particularly nerdy and excited about and I have to share it with the Archives Department.
We need an example of that. Get nerdy with us.
OK, you're putting me on the spot. So he was heading into the Christmas holidays of 1899. And he had come up with a list of students who were worthless and not worthless-- or hopeless, I'm sorry, hopeless. Hopeless.
Wow.
In his mind, I think, based on reading all these letters for-- I've read now like 3,500 letters of his-- and hopeless, it's I've given them an opportunity multiple times to show me that they're being responsible students and they're not. So in his mind, this is my hopeless list and this is my not hopeless list.
And he sent letters to the parents during the holidays, letting them know that their student may need to consider not coming back to Georgia Tech, which you can imagine, during the Christmas holidays, now, his outgoing correspondence changed to a lot of fielding angry letters from parents saying, "you're kicking my kid out of Georgia Tech?" It's funny and interesting.
I think it would make national news if a university president called students worthless. KATIE GENTILELLO: I think so, too. Yeah. Yeah. CHARLIE BENNETT: Alex, these letters, are they an example of the content is a lot more important than the object itself?
Yeah, I would say the information in the letters is what is valuable as opposed to the physical item itself. Because preserving onion skin is a losing game for us. And so that's why this is a really great example of why-- this is an ideal digitization project. We are rushing to save the content because of how rich it is. And now we'll have these digital scans. That is the primary use copy we will have now. And that's our preservation copy now, too, almost.
You are listening to Lost in the Stacks. And we'll talk more about archives and digitization on the left side of the hour.
[MUSIC PLAYING]
Hi, I'm Kim Stanley Robinson, science fiction writer and a fan of The Ramblin' Recks of Georgia Tech. This is WREK. [MUSIC PLAYING] And I know And I know And I know that everybody And I know that everybody be happy It's happy as you and me 'Cause I know.
This is Lost in the Stacks. And our show today is called "You Can't Digitize Everything." It's the first in a new occasional series, "Recalling Stories That Archives Tell," which is shorthand for the unwieldy original title. Let's consider what we've put into archives in order to craft the rough draft of histories both local and global. And remember, that if you find something in an archive, it's because an archivist put it there, not some random event.
And also, let's stop imagining huge single-action solutions for the ever-changing problems of access, preservation and discovery in our cultural institutions. And I thank Alex for joining the team and getting a shorter title there. I want to tell you all a quick story. On May 10th, 2009, Computerworld published a piece by columnist Mike Elgan titled, "Why You Should Digitize Everything." So I had to read that, right? Here's how it starts.
"Two events this week, one personal and another that is making international headlines, made me rethink what can and should be digitized-- everything." From here, Mr. Elgan says that his work as a writer and the nature of the internet at the time meant that he no longer needed to be in the same house year round. So he and his family were going to become semi-nomadic, spending most of the year traveling. This meant selling the house, downsizing dramatically. That's his personal event.
The international event was raging wildfires in Central California, which made it unflinchingly clear that any collection of physical things that his family kept was in danger of destruction. His proposed solution was to get rid of everything, and whatever was somehow important and irreplaceable they would digitize. He wrote, "When it comes to deciding whether to keep or discard something, where do you draw the line? Old holiday and birthday cards? OK, those can be discarded.
Mother's Day cards from kids? Hmm. Trophies? Yikes. There are a million items that make you feel a loss when you toss. But if you keep them, they'll be buried unseen for decades." His solution was a five-step process. One-- capture. Take digital pictures of everything. Two-- index for search, and let me quote him at length on that. "Sign up for an account with Evernote," he says. Download the desktop application. And I should say, members of the show are now laughing.
Download the desktop application and drop all your pictures into the application. Evernote will upload them all to its servers. And here's the best part-- index all words it finds in the pictures, which makes them searchable. Later, you can just search Evernote as if it were Google. I'm going to let that one go by. And find pictures of just about any item. You can also categorize, tag, sort, or file everything in any way you choose. Then he says, "Step three is share.
Upload sentimental and appealing pictures to social media accounts. Four-- backup." And he suggests a paid service. "And finally, number five-- discard." As he says, "Here's the best part. Shred, recycle, burn, or discard most of this stuff you digitized. We no longer have to manage all those boxes. Best of all, neither fire nor time nor neglect can destroy the photographs we took of them." OK. Here is a post by a man named Mark Foerster on an Evernote forum on March 10 of this year.
Quote, "Evernote has now forced me to change from the legacy version, i.e. the version that actually worked, to Evernote 10. Now that I have been forced to leave Legacy, I found that 10 is even worse than it was when I first tried it. In fact, it's an absolute nightmare. Nothing in the right place, nothing is intuitive. It's very ugly, difficult to use, and missing some of the features which I used a lot, for example, presentation mode. I've been with Evernote from its very earliest days.
I have got a vast amount of stuff in it. Moving would be a major effort, yet I think I'm going to have to. What are the latest recommendations for alternatives?" I hope everyone can draw their own conclusions. And we can file this set under PR6037.C23.
[DAVID BYRNE AND BRIAN ENO, "STRANGE OVERTONES"]
That was "Strange Overtones" by David Byrne and Brian Eno, a song about figuring out how to achieve a goal.
[MUSIC PLAYING]
This is Lost in the Stacks. And our show today is called "Stories Archives Tell-- You Can't Digitize Everything," the Lyman Hall edition. Our guest is Katie Gentilello, Digital Projects Coordinator at the Georgia Tech Library. So Katie, you said in the last segment that you've read a lot of these Lyman Hall letters. Do you have a number in your head, how many you've read? KATIE GENTILELLO: Probably about 3,500. 3,500. I don't even know how to start. What's good about them? What's in there?
What have you learned reading 3,500 letters from an ex-president of Georgia Tech?
I did a presentation to the department about a year ago about some of the things that I've uncovered. I would say, I introduced Lyman Hall as my friend Lyman Hall. Because when you step back into personal letters like that, even in the position that he was in as a president, you get a sense of who that person was, what's important to them, what isn't important. We know that he had an untimely death while he was the president of Georgia Tech, likely due to the stress.
And you can get a sense of that stress that he's increasingly underfundraising and dealing with an increasing number of matriculated students. They started out at 150. In 1900, they're up to 400. They're asking for more dorms. They're asking for more resources from the legislature that are not being forthcoming. He's asking his friends and benefactors for more money. He's asking for, with the textile school coming online, he's asking for local cotton mills to donate materials.
He's not getting much of a response. So you can see that he's in a hopeful position, but he's also dealing with the large amount of stress of just trying to get this institution up off the ground. He mentions overtures to go and be the president of other schools and turns it down because he has strong feelings about getting Georgia Tech up and being the preeminent engineering school in the South. He's very proud of that. And you sense it.
You sense just how convicted he is about the success of the school. It's a really neat kind of like pulling the wind, opening the window and seeing what early Georgia Tech was facing in terms of its success and just how much blood, sweat, and tears went into getting the institution up off the ground.
And if anyone tried to do this project again, these letters would start falling apart in their hands, right? KATIE GENTILELLO: Oh, well, they-- Yeah. They are falling apart even as we scanned them. Yes, unfortunately. How many have you lost, do you think?
Well, the volumes themselves are intact. The individual pages are brittle and are falling apart. I can think of 2 of the 10 volumes right now that are in critical shape, to the extent that they probably need to just be closed and restricted from now on.
And I think we made that call earlier this year, is that we will not provide access to those physical volumes anymore. Because with the scans, there's just no reason for us to do that, to jeopardize them. And people freak out when you start having flakes of pages fall apart anyway. So we can all just spare everyone some stress. CHARLIE BENNETT: Katie, what have you learned about digitizing doing this project maybe that you didn't know before? Or did you learn a new process?
Or did you figure out how to avoid an old mistake? KATIE GENTILELLO: Yeah, so this is the first project where I've had to manually transcribe this many words. It's always a decision, a very weighty decision at the front end when you have to decide is this collection useful in its current digital form. And by that, I meant just scanning the pages, is that enough? And I realized that it wasn't And I tried multiple ways of-- I tried OCRing.
I tried different techniques to try to make that happen programmatically and it just was not successful. And so to make the decision to have to manually transcribe them was a very big decision to do it. And before I set out asking my student assistants to do it, I did a few pages myself just to time it, just to see how difficult it was going to be.
And once I figured out a Photoshop filtering strategy for the derivative files-- I have the original color files and then I make them black and white and punch up the contrast-- once I was able to get those contrasted, digital surrogates created, it made reading the letters much easier. And so that was a process I hadn't had to do before. And then hand-transcribing is definitely something that-- nobody wants to undertake this many letters doing it by hand.
But everyone that's had some involvement in it has really enjoyed uncovering the information in it. And I'd say the bonus is that, yes, it's a lot of work. But the knowledge that just myself is gleaning from having to read every single page for quality control checking is kind of making me an expert on Lyman Hall. And just case in point, when we started digitizing this collection, the Fulton Bag and Cotton Mill exhibit was getting ready to come online.
And the timing of the digitization of this collection and the exhibit coincided to the point where we were able to make connections from Georgia Tech back to Oscar Ellis and Fulton Bag and Cotton Mill and what their relationship was that had not been able to be known before. So we were able to connect the dots and use some of the letters from that collection to support the narrative in the exhibit. So just being able to uncover these little breadcrumbs like this has been really, really fun.
So Alex, is Katie now your subject matter expert for Lyman Hall presidency?
Oh yeah. For Lyman Hall. Yeah.
I wasn't kidding, he's my friend. My friend, Lyman Hall.
You probably know more about Lyman Hall than anyone else in the country after reading 3,500 letters.
Yeah. Yeah. I'll know more because there's 10,000 letters total.
Oh my gosh.
Yeah. So we're about a third of the way through.
Alex, as university archivist, can you speak a little bit to why is this a good thing to do?
Obviously, as Katie has already talked about, there's a lot of really critical information about how Georgia Tech was developing in the early years. The first president of Georgia Tech, that was just like opening the doors. Lyman Hall really pushed both our academics, our growth, our campus development in a way that we haven't seen quite yet. And with his ability to fundraise, that was something that was very new for Georgia Tech. And it really, very rapidly changed campus.
So his presence in Georgia Tech history is really notable. And so to think that there was a preservation concern where we're worried about losing this information, that is then another layer of this is even more critical that we preserve this really unique, significant piece of our institute's history. And to have it digitized, obviously, the access component of it cannot be understated. I mean, there's a reason why everyone wants to digitize things.
It's because it makes things immediately more accessible, more usable. There aren't boundaries in terms of having to come into our reading room and view stuff. So it's really exciting that we're going to have this online. And anyone interested in learning about these very critical years for Georgia Tech's formation will be able to access it and view it and delve in and become maybe another Lyman Hall expert in addition to Katie.
This is Lost in the Stacks. And today we've been talking with Katie Gentilello, Digital Projects Coordinator at the Georgia Tech Library, about the good, the bad, and the ugly of digitizing everything. Katie, thanks for joining us.
Thank you for having me. This was fun.
And you can file this set under BF449.5.C38.
[THE BREEDERS, "WHEN I WAS A PAINTER"]
[JESUS LIZARD, "HIDE AND SEEK"] Hide and seek Hide and seek That was "Hide and Seek" by Jesus Lizard from the new record, first in 26 years. And before that, "When I Was a Painter" by The Breeders, songs about reflecting on the past and planning for the future.
[TELEVISION, "FRICTION"]
Today's show was called "Stories that Archives Tell-- You Can't Digitize Everything," the Lyman Hall edition. In the time we have left, I wonder if we can all be candid about whether we each personally want to digitize everything or not. So I put it to you first, Alex. Do you want to digitize everything?
No. Hard no.
[LAUGHTER]
There's some things that are not meant to be digitized.
Nice. Fred?
There are some parts of me that want everything that is digital to disappear one day. That thought gives me a lot of comfort. But then again, it would be nice to be best friends with someone that lived 120 years ago. MARLEE GIVENS: That's a good point, but I'm going to say no as well. And part of it is because, that you now have a new thing to take care of in addition to the old thing.
Katie.
Well, I think you know my answer. I'm in this business and I love digitizing. I love the idea of it. Nothing can beat the feeling of an actual artifact of a photograph, a book. The physical materials that you can hold in your hand, there's something very special about that. And that's something I would never want to lose. CHARLIE BENNETT: I have to confess, and I feel like I'm not the one who should be thinking this, but I have recently wanted everything digitized.
I've been wanting to be able to look back, find answers to things. I've been looking at property records from 25 years ago. I'm not going to tell you why. But I've really wished that I could just find them instead of, I think, having to go to a basement somewhere. But I know that we shouldn't. So let's just roll the credits.
[SOUNDGARDEN, "FOPP"]
Lost in the Stacks is a collaboration between WREK Atlanta and the Georgia Tech Library, written and produced by Alex McGee, Charlie Bennett, Fred Rascoe, and Marlee Givens.
Legal counsel and letters about being utterly hopeless provided by the Burrus Intellectual Property Law Group in Atlanta, Georgia.
Special Thanks to Katie for being on the show, to every archivist who has patiently explained why something isn't on the web, to all the radio listeners who heard a version of this show already. And thanks, as always, to each and every one of you for listening.
Our web page is library.gatech.e du/lostinthestacks, where you'll find our most recent episode, a link to our podcast feed, and a web form if you want to get in touch with us. CHARLIE BENNETT: Next week, I don't know what's going to happen. If you've been listening to Georgia Tech News, you know that the whole campus shut down for a little bit and maybe it's coming back and maybe it's not. Let's just see what goes on next week, OK, everybody?
It can all disappear. That's what I'm telling you. It's time for our last song today. I'm imagining all those pages in the collection we talked about today, 10,000 or so waiting, just waiting, to be handled. All projects have steps that you need to take and repetitive actions. But this one in particular feels like everything is in line waiting to be processed. All those items waiting to be digitized like a long line of cars trying to break free.
So let's close with "Long Line of Cars" by Cake, here on Lost in the Stacks. Have a great weekend, everybody.
[MUSIC PLAYING]