Summary
Computers have found their way into virtually every area of human endeavor, and archaeology is no exception. To aid his students in their exploration of digital archaeology Shawn Graham helped to create an online, digital textbook with accompanying interactive notebooks. In this episode he explains how computational practices are being applied to archaeological research, how the Online Digital Archaeology Textbook was created, and how you can use it to get involved in this fascinating area of research.
Introduction
- Hello and welcome to Podcast.__init__, the podcast about Python and the people who make it great.
- When you’re ready to launch your next app or want to try a project you hear about on the show, you’ll need somewhere to deploy it, so take a look at our friends over at Linode. With 200 Gbit/s private networking, scalable shared block storage, node balancers, and a 40 Gbit/s public network, all controlled by a brand new API you’ve got everything you need to scale up. Go to pythonpodcast.com/linode to get a $20 credit and launch a new server in under a minute. And don’t forget to thank them for their continued support of this show!
- And to keep track of how your team is progressing on building new features and squashing bugs, you need a project management system designed by software engineers, for software engineers. Clubhouse lets you craft a workflow that fits your style, including per-team tasks, cross-project epics, a large suite of pre-built integrations, and a simple API for crafting your own. Podcast.__init__ listeners get 2 months free on any plan by going to pythonpodcast.com/clubhouse today and signing up for a trial.
- Visit the site to subscribe to the show, sign up for the newsletter, and read the show notes. And if you have any questions, comments, or suggestions I would love to hear them. You can reach me on Twitter at @Podcast__init__ or email hosts@podcastinit.com)
- To help other people find the show please leave a review on iTunes, or Google Play Music, tell your friends and co-workers, and share it on social media.
- Join the community in the new Zulip chat workspace at pythonpodcast.com/chat
- Your host as usual is Tobias Macey and today I’m interviewing Shawn Graham about his work on the Online Digital Archaeology Textbook
Interview
- Introductions
- How did you get introduced to Python?
- Can you start by explaining what digital archaeology is?
- To facilitate your teaching you have collaborated on the O-DATE textbook and associated Jupyter notebooks. Can you describe what that resource covers and how the project got started?
- What have you found to be the most critical lessons for your students to help them be effective archaeologists?
- What are the most useful aspects of leveraging computational techniques in an archaeological context?
- Can you describe some of the sources and formats of data that would commonly be encountered by digital archaeologists?
- The notebooks that accompany the text have a mixture of R and Python code. What are your personal guidelines for when to use each language?
- How have the skills and tools of software engineering influenced your views and approach to research and education in the realm of archaeology?
- What are some of the most novel or engaging ways that you have seen computers applied to the field of archaeology?
- What are your goals and aspirations for the O-DATE project?
Keep In Touch
- Blog
- @electricarchaeo on Twitter
Picks
Links
- O-DATE Textbook
- Carleton University
- Ottawa Canada
- Simulation Modeling
- Agent Based Modeling
- NetLogo
- Complexity Theory
- Archaeology
- Digital Archaeology
- The Programming Historian
- University of Western Ontario
- Historical GIS
- ArcGIS
- QGIS
- Digital Humanities
- Project Jupyter
- Binder – Service for hosting Jupyter notebooks
- E-Campus Ontario
- Graph Databases
- SparQL
- OpenContext.org
- TDAR (The Digital Archaeology Record)
- R Language
- R OpenSci
- Arrow
- Pandas
- Neural Networks
- Generative Adversarial Networks
- Computer Vision
- Archaeogaming
- Alamagordo Atari Excavation
- Leiden University
- Interactive Pasts Conference
- Photogrammetry
- LIDAR
- Palmyran Arch
- Ben Marwick
- Matt Harris
- Jolene Smith
- Sara Perry
- Rachel Opitz
- Colleen Morgan
- Patrick Burns
- Ethan Watrall
- Andrew Reinhard
- Neha Gupta
- Katherine Cook
- Value Foundation
The intro and outro music is from Requiem for a Fish The Freak Fandango Orchestra / CC BY-SA
[00:00:00]
Unknown:
Today, I'm interviewing Sean Graham about his work on the Eimeline Digital Archaeology textbook. Hello, and welcome to podcast dot in it, the podcast about Python and the people who make it great. When you're ready to launch your next app or want to try a project you hear about on the show, you'll need somewhere to deploy it. So say hi to our friends over at Linode. With 200 gigabit private networking, scalable shared block storage, node balancers, and a 40 gigabit public network, all controlled by a brand new API, you've got everything you need to scale up. Go to pythonpodcast.com/linode, l I n o d e, to get a $20 credit and launch a new server in under a minute. And don't forget to thank them for their continued support of this show.
And if you're like me, then you need a simple and easy to use tool to keep track of all of your projects. Some project management platforms are too flexible, leading to confusion of workflows and days' worth of setup, and others are so minimal that they aren't worth the cost of admission. After using Clubhouse for a few days, I was impressed by the intuitive flow. Going from adding the various projects that I work on to defining the high level epics that I need to stay on top of and creating the various tasks that need to happen only took a few minutes. I was also pleased by the presence of subtasks, seamless navigation, and the ability to create issue and bug templates to ensure that you never miss capturing essential details.
Listeners of this show will get a full 2 months for free on any plan when you sign up at python podcast.com/clubhouse. So help support the show and help yourself get organized today. And don't forget to visit the site at pythonpodcast.com to subscribe to the show, sign up for the mailing list, and read the show notes. And don't forget to keep the conversation going at python podcast.com/chat. Registration for PyCon US, the largest annual gathering across the community, is open now. So don't forget to get your ticket, and I'll see you there. Your host as usual is Tobias Macy. And today, I'm interviewing Sean Graham about his work on the online digital archaeology text book. So, Sean, could you start by introducing yourself?
[00:02:10] Unknown:
Hi. I'm Sean Graham. I'm an associate professor of digital humanities in the history department at Carleton University in Ottawa, Canada.
[00:02:18] Unknown:
And do you remember how you first got introduced to Python?
[00:02:21] Unknown:
I I think what happened was when I first started at Carlton, I my my computational background was in simulation modeling in agent based modeling. And I I'm not particularly good at any of this stuff. And I was using a platform called Metlogo, which was originally developed to teach high school students about complex systems and complexity theory and and that sort of thing. And when I joined the history department as an archaeologist, I felt like a bit of an imposter, and I needed to quickly learn everything that was going on in other facets of history and archaeology and to figure out how I could fit into everything that was going on here. And a little bit of googling around later, I came across version 1 of the programming historian, which was an online textbook in series of tutorials by Bill Turkel at Western at the University of Western Ontario and Adam Krimbill.
And he had a sequence of common tasks that a historian might want to do programmatically, and they were all written in Python. So I sat down and worked my way through those, and it's been, you know, how fast can I search for something on Stack Overflow? But it was very much courtesy of of Bill and Adam's original work, which came out in about 2, 008, I think. So I started here in 2010.
[00:04:01] Unknown:
Well, it's, opportune happenstance that Python has continued to grow in popularity and use, which I'm sure has simplified some of your ongoing efforts to apply programming practices to your work in the archaeology department and or, sorry, in the history department as an archaeologist?
[00:04:20] Unknown:
Oh, yeah. Very much. And it's it's interesting because everything you know, there are fashions in everything, and there are fashions in history as well as in archaeology. And there there's been a lot of interest in in mapping and web mapping and historical GIS. And things like ArcGIS have rather expensive licenses. So open GIS software, things like QGIS, which has a lot of modularity built into it and and can be scripted in Python and so on, means that trying to stay 1 step ahead of where my students are interested keeps pushing me to to learn new things.
[00:05:04] Unknown:
Yeah. I've had occasion to use QGIS myself for some physical installations of IoT sensors in a past job, and it was definitely useful for being able to have a open and accessible platform to tinker with and build up maps of where the different sensors were and try to create overlays of what the, network paths were.
[00:05:27] Unknown:
Absolutely. And the the intersection of Python with all sorts of different open source platforms has been a real boom for me because I can't I'm not allowed to, assign a particular software package as a required purchase to students. I can I can put down, you know, purchase this $200 textbook without issue? But if I say you you have to buy a $20 subscription to something, that I'm not allowed to do. So we we need to do some catching up. But in the meantime, it pushes me to use a lot of open source stuff, which is great because then the students learn how to fend for themselves and figure out how to find the packages or the software or the platforms or whatever that can help them reach their research goals.
[00:06:17] Unknown:
1 of the few cases where the unintended side effects of bureaucracy ends up being positive. Absolutely. Yeah. And so as part of your work as a professor in archeology and in the history department, you have collaborated on this online digital archaeology textbook or Odate for short. So can you start by giving a bit of an explanation of at least how you see the definition of digital archaeology?
[00:06:47] Unknown:
Sure. For me, digital archaeology is not just about using computational power to ask archeological questions or to answer archeological questions. It also involves thinking through what these tools and approaches do to how I understand or conceive the past as well. So in the Open Digital Archaeology textbook, we have a lot of support, I suppose, for using digital tools to deform our archaeological material or deform our perspective on what we're looking at, to think through creative uses, so computational creativity as well as straightforward, you know, spatial statistics or or what have you. So it sits in a broader digital humanities framework.
Archaeology has been using things like GIS for 40 years and there's any amount of support for teaching GIS, but there's much less support out there for trying to say you're interested in how visitor impacts on an archaeological site in a national park and maybe what you wanna be doing is scraping Instagram and studying comments for what people are attending to and what what they find is important or or meaningful. So that's the kind of thing you can also learn courtesy of of our textbook. And in teaching or trying to teach these kinds of methods in my own classes, I end up wasting 2, 3 weeks on tech support and getting everybody on the same page.
I have to rely very much on my students bringing their own device, but I can't be supporting 3 different flavors of Windows and machines that might have 2 gigs of RAM and or 7 gigs or or the odd person with a Linux box or, you know, Mac machines from 4 or 5 years of of of different specs. So with Odate, we also the the original vision was to have an entire virtual machine that could be installed with 1 click, and the textbook would be inside the virtual machine as well. That proved to be really difficult because of the security implications and the university not wanting to open ports to to anybody to be running code on our servers.
I was developing all of this and proposing all of this right in the middle of a ransomware attack at our university. So, you can imagine how nervous what I was proposing was was making people. But it was around about that time that Tim Sherritt, who is a digital historian in Australia, sent me a direct message on Twitter saying, have you seen this binder service? And that that really enabled everything to take right off because I could just package up targeted exercises or packages or or, Jupyter Notebooks in a single repo, 1 click, boom, binder, virtual environment, and we were off to the races.
So it's, it it was very much different what I ended up building than what I initially proposed. And this was funded by an outfit called Ecampus Ontario, which is a an agency of the Ontario Provincial Government that wanted to, wants to support Open Educational Resources. And in their initial, description of the call for proposals it was clear that they imagined digital textbooks as being basically PDFs that a person could download and and maybe print out and put into a course pack. So an online digital textbook with Jupyter notebooks and binder powered environments, really pushed the envelope for what they were imagining.
[00:11:07] Unknown:
And in the end result, it ends up being much easier to keep things relevant and up to date as well because you don't have this static asset of a PDF, for instance, that somebody might download and can quickly have issues with various different versions floating around or not knowing how up to date something is, or if new versions of packages come out and you're having source code embedded within the PDF that might not actually be accurate anymore versus this binder environment that is hosted, that is always going to be consistent for everybody who accesses it within a given time frame. So I imagine that has greatly simplified your process of teaching these systems as well. Oh, yeah. Absolutely.
[00:11:54] Unknown:
And and it's great too because, you know, when we revealed this to the world, I said, and I identified Twitter and via my blog, you know, I said, look, this is there's gonna be rough edges. There's gonna be things that don't work. There's gonna be things that get dated or packages are gonna change. That's okay. That's part of the learning. That's part of what we want you all to understand. Here's how you can contribute. Here's how you can suggest changes. Here's how you can tell us when things are broken, this is how we can fix it. And so it becomes becomes very much a living document. And we've already had some poll requests from people external to the project improving, our section on databases, for instance, adding adding sections on graph databases.
But even 1 of my even 1 of the Jupyter Notebooks that we built for our initial launch, it was, it was using or it was trying to teach about principles of linked open data and how to do sparkle queries and so on. On. And we were using the endpoint from the British Museum. The British Museum totally pulled the plug on all of that the people who'd been supporting it and and championing it at, at their institution. And so that endpoint went down and is no longer supported. So, you know, things change quickly in in the the world of museums and and, cultural institutions as far as their their online stuff goes.
So if if we were building things in a static sense, that would be an enormous disaster for us. But because we anticipate, we expect this kind of stuff to happen in the first place, the totally, the repos and notebooks and binders, turns it into a a learning experience for everybody. So it's it's authentic. It's real.
[00:13:45] Unknown:
And 1 thing that I found interesting in the process of reading about the textbook and some of your blog posts on the matter and looking through the material is the concept of a perpetual beta that you are supporting for this work, which means that there as you said, you decided to go ahead and launch it, even though it wasn't necessarily perfect in your mind, but recognizing that it will never reach that state of perfection, especially if you never release it to the world. So wondering if you can talk a bit about your philosophies of how you view the textbook as this living entity and some of your hopes for how the community will engage with it both in terms of the software community and the broader academic community?
[00:14:33] Unknown:
Yeah. Well, the part of the advantage of a perpetual beta framework is I'm I'm not expert in every single thing that we've covered in the in the volume, and I never could be. I have I have deep experience in a extremely narrow aspect of of all of this. So I I don't think we could ever reach the kind of perfection that you would want for or that we have been trained to expect for print publications. Right? You you put something out in prints and you never revisit it. By by focusing it or branding it, maybe that's a better word, but by saying to people, look, it's perpetual beta means that my hope is that when people encounter things that are a bit wonky or not as clearly explained as we might want, that they'll take it in the spirit in which it is offered and try to to improve things.
And in that sense, if people fork it and it ends up going in in many different directions, that's that's good. I I don't wanna be prescriptive about what digital archaeology is or could be, but rather the people who who need it or who want it and take it forward become in the driver's seat. And I know that it's it's gonna mean that I have to always keep an eye on what's going on, and, eventually, maybe I'll be able to handle the the official reins to somebody else. But it also opens up teaching opportunities for me too with my my graduate students, my undergraduates, about what it means to to build community or to to try and keep something like this as current as possible.
So Perpetual Beta, in 1 sense it's a bit me covering my ass and recognizing that I couldn't hope to put out something that covers all the bases and is perfect the first time. But I, more seriously, I I regard it as an open invitation, and I hope that people will, you know, will accept it in the in the spirit in which it is offered.
[00:16:52] Unknown:
And in terms of the subject matter that you cover within the textbooks and the notebooks that go with it, I'm curious what you have found to be some of the most critical lessons for your students in order to help them be effective and successful archaeologists?
[00:17:09] Unknown:
The, the the idea about failing productively and and how to to recover from things that don't work or things that collapse on you. My my students coming at me in the history department and to a lesser degree in archaeology, they they have been trained for years to just accept what the computer gives them and to open the hood and try and get it to do something else to, you know, the kind of things that we, you know, growing up in the eighties, you sit down at a Commodore 64 and you turn it on and it says ready and it just blinks and it does nothing unless you start telling it what to do. The my students, at the very least, come with a huge degree of anxiety about about trying anything other than what the software already gives them. So they very much live inside the boxes that are are provided.
Teaching people that it is okay to open the hood, that it's okay to take control of of your machine and to push back against the decisions that other people have made in their code so that you can do your research more honestly or more effectively or in new directions. That's that's a really hard thing to teach. So in the in the textbook, we put a lot of effort into talking about a productive fail mindsets, about open and reproducible code, about sharing your code for replicability and reproducibility. These are things that we don't normally teach in at least on the humanity side of the, of the quad.
So the tech will always be changing. But if you've got that good mindset from the outset, you'll be able to figure out what it is you need out of the tech and where to find answers and how to roll with it. And that, I think, is probably the single most useful thing that comes out of this whole experience.
[00:19:22] Unknown:
In terms of their experience once they graduate and get involved in actually doing archaeological research or field work, I'm wondering what you view as being the level of necessity of that computational comfort and experience as they progress throughout their careers?
[00:19:41] Unknown:
I think it's only gotta be to the good. If we can teach people how to figure out what they need to know, when they need to know it, and where to go in order to get the solutions. I mean, I spend huge amounts of time on Stack Overflow. Thank you to those folks. Then just getting in the habits of thought and understanding how code works and how to read it and and what to where to troubleshoot. Like, those are the things that, to my mind, are are more valuable than experience with any particular package at a time. Right? Like, it's it's the problem of of GUI based software packages in the first place. If the the GUI changes, then you don't know what to do anymore.
So does does that answer that question?
[00:20:33] Unknown:
It does in part, but I was also curious in the broader sense of the amount of crossover of the computational tools that are required for doing archaeological research and field exploration for people who are engaged in that in a professional context both now and as you see it progressing into the future?
[00:20:56] Unknown:
Crossover from what we're teaching to their professional lives?
[00:21:00] Unknown:
Just in terms of archaeological research, I'm just curious, like, how prevalent using software packages and computational techniques is applied versus spending time digging up artifacts or researching physical structures, things like that. Right. Right.
[00:21:19] Unknown:
Well, in in commercial archaeology right now and I and I mean I have to be honest, I haven't worked in commercial archaeology in quite some time. I actually have very little experience in commercial work but there's computational work happening all the time Databases, mapping, GIS, public outreach for all of it. I'd say the the computational work is more more the bread and butter than the the actual field work as it were because as the field work happens, the the material is being digitized as the field work goes on. Right? Things are being excavated, things get recorded into various databases, the the whole analytical side that happens the field work is you know maybe 1 third of what goes on and it's all the behind the scenes stuff, the, the assessments that happen before field work goes on, the predictive work, the the archival work to, work out the probabilities of what kinds of materials you might reasonably expect. Like, that sort of work is very computational and probably occupies far more of the the working life of of people in commercial work than than than field work does.
[00:22:47] Unknown:
And in terms of the data sources and data formats that would commonly be encountered, you mentioned some of these geographical surveys and GIS information, digitization of some of the artifacts that might be recovered, whatever formats that might take, and then also recording of metadata. And, I'm sure there are time aspects to when the artifacts are recorded or when the observations are taken. So I'm just wondering if you can talk through those different types of data and the sources and locations that they might be found for people who are engaged in performing these computational analysis of all of that information.
[00:23:29] Unknown:
Right. Well, it's, it's it's a god awful mess. There is material trapped in image based PDFs. There's stuff in old versions of Excel. There are XML, JSON, CSV, FileMaker, God help us all, Word documents. There there's there's no there there are archaeological repositories for digital work, and it's not necessary not necessarily true that every academic archaeologist or every commercial archaeologist is putting materials into these repositories. Some of the bigger ones are open context, open context dot org, which tends to have everything represented in JSON. Then there's TIDAR, which is out of Arizona State University, I believe, and they are more of a they they will take your documentation in all sorts of different formats, whereas OpenContacts works with you as a data publisher to get your data into a format that can be exposed through their API and can be recombined and reused with other stuff. There's the archaeological data service in the UK, which again will take materials in lots of different formats, everything from field notes to to tables.
So you you have you know, 1 of the things that we, 1 of the binders that we have in Ode uses an R package called Tabulizer to, to try and, when you encounter data locked in a PDF to to get it into a data frame that you can then use. So there's there's some folks are using graph databases. Some people are using, all sorts of different schemas. It's, you you can you can encounter data in all sorts of different formats. A friend of mine who's, I forget what the the proper title is, but it is states a state archaeologist responsible for a lot of the data management of things. She encounters material kept in word files, material kept in old shoe boxes. It's it's a wild west of data.
[00:25:59] Unknown:
And in terms of the notebooks that, you and your collaborators have created to accompany the textbook, there's a fairly even mix year of R and Python code. So I'm wondering what you have found to be your personal guidelines for when to use each language to its greatest benefit.
[00:26:20] Unknown:
I'm not sure that I have anything really useful to say on that because I'm I'm a bit of a magpie when it comes to to code. If if I can get it to work and it does roughly what it is I need it to do, then that ends up being the the language or approach that I I end up using. So I got into using R via text analysis and topic modeling and that sort of thing. A lot of archaeologists are more comfortable with Python than they are with R. That said, Ben Marwick at, Washington State University has a huge compilation of R packages that are useful for archaeology. Matt Harris, is another 1 who's very, very handy with our can do all sorts of interesting things with it.
So, really, if it does the job you needed to do, then that's the package you use. Yeah. Not the most helpful approach, I suppose. But as long as your code is is well commented out and you understand what it's doing, then either or, I suppose.
[00:27:33] Unknown:
Yeah. And I think that just in the broader sense of software, once you have enough of a foundational understanding of what's happening, then the syntax to some degree starts to melt away, and it's much better to be pragmatic than dogmatic in your approach to when to use which tool.
[00:27:52] Unknown:
Absolutely. Absolutely. I'm in I'm entirely self taught. I, I screw up and make errors all the time. So I really appreciate people who put out packages that are are cleanly explained and and commented and have examples that I can work through and think, okay. I see it. This is how their data is organized. I can get my data organized like that. I can use their sample code. Now what can I plug this in with to get to the next stage that I need it to be at? So the the IR OpenSci folks, they they're they've got lots of really great packages, and and I really appreciate what they're up to as well.
[00:28:34] Unknown:
Yeah. And particularly in recent years, there has been a lot more work done to ease the interoperability between R and Python in the data analytics space with things like the arrow project to make data frames interoperable between R and Pandas, and just a a general approach within the community of, academics and scientists and data scientists to lower the friction and just make collaboration easier regardless of the language community?
[00:29:07] Unknown:
Absolutely. And I'm I haven't I haven't explored that project very much yet, but that's that kind of work is is really important, but it still requires an interpreter to get it down to the level of where so many of our students are. So that's something I'm gonna have to put on the agenda for future updates to, to Odate.
[00:29:30] Unknown:
And in terms of your overall experience and career, I'm curious how the skills and tools and thought models of software engineering have influenced your views and approaches to the way that you conduct research and education in in the realm of archaeology and history?
[00:29:51] Unknown:
Well, I I feel in some senses, I feel like I'm 1 step removed because I'm I'm very much influenced by people like Carl Boetticher, and I'm not entirely sure I pronounced that correctly, who was big on open notebook science and and and Ben Marwick, who's an archaeologist, or Caleb McDaniel, who's a historian of these these ideas of replicable code chunks and and being open about our processes entirely as academics. So, the degree to which the actual software community influences. I'm sure it does. I I'm sure it is because I'm, you know, I'm picking over carefully through people's blogs and Stack Overflow posts and trying to figure out my own immediate concerns.
And I expect I I'm picking it up through osmosis, but I'm not I don't know that I could point to any single individual or or group of of people.
[00:31:00] Unknown:
And in terms of the ways that, software and analysis and computational capacity are applied to the field of archaeology. What have you found to be some of the most novel or engaging ways to make use of that?
[00:31:20] Unknown:
Well, I'm really interested in neural networks, for computer vision at the moment, and I I'm fascinated by things like generative adversarial networks and people exploring the the kind of latent space of images. And I'm I I have a research project on at the moment that looks at people who buy and sell human remains and trade them on, via social media and using computer vision to try and work out some of the rhetorical tropes of the visual composition of these images. Like, what is it that people are how are they signaling that things are for sale or what kinds of ideas that have you know, 1 1 thing that we're starting to notice is how when people are trying to signal something's for sale, but not actually stayed as much in terms of the comments, the the images end up being composed rather like how we would imagine a museum would display human remains of the, like, the 19 thirties or forties.
So sort of an Indiana Jones visual motif, coming through in the images, which, you know, you might pick up on if you were staring at pictures all day. But when we do it at scale with convolutional neural network, it it stands out an an awful lot more. So that kind of thing is, you know, where I'm at at the moment. Other people are using them to pick out features in, field photographs. So as you're excavating you're taking photos of each level or feature or context as you dig down. So every event that happens in the formation of a site is a context and you can peel these back like layers of onion or and the way that they intersect is a sort of three-dimensional, 4 dimensional puzzle.
And there are some folks who are feeding 100 or 1000 of these kinds of photographs into neural networks to see what, what can we pick up when we let the computer do the looking rather than us. So, you know, interesting things like that.
[00:33:47] Unknown:
And another piece that I saw referenced as I was preparing for this show is the concept of archeogaming. So I'm wondering if you can talk a bit about what that is and some of the ways that it's applied.
[00:34:00] Unknown:
Right. So so archeogaming as a term was coined by Andrew Reinhart, who is a classical archaeologist and, archaeological publisher. And he was also involved with Bill Caraher and a few others in the Alamogordo excavations of the Atari burial ground. So if you've seen the game over documentary on Netflix about those excavations, you will have seen Andrew Reinhart. And he's actually doing a PhD in this at the moment, which is quite exciting. And he and I and a number of others for a couple of years have been exploring ideas of what does it mean for humans when so much of our culture these days is digitally mediated or takes place within digital spaces.
And archaeology is about the material expression of human culture whether it's in the remote past or the current current day. So if we do archaeology inside these digital spaces, what does that look like? How do we translate the methods and perspectives of dealing with material culture into, into digital culture? So archaeogaming can be the application of archaeological perspectives within constructed digital spaces. It can be looking at the representation of human past or material culture through video games. So 1 aspect of archeo gaming also intersects with archeological ethics when you think about loot looting mechanics and loot boxes and, archaeological tropes within video games.
There's a there's another group at the University of Leiden who have been doing a lot of work in in ArQube Gaming and they run an annual conference called the, the interactive pasts conference. So it's, it's kind of an exciting field. Right? Because it's a new new avenue for for human, for archaeology to try and understand the human condition. Yeah. Humans have been making virtual worlds since the the Neolithic. Right? You look at the caves of Lascaux, you look at, you know, all the way up to Disneyland, that's another virtual world. So now it's moving into these, crafted virtual worlds that we interact with through the screen.
[00:36:25] Unknown:
Yeah. It's definitely a very fascinating area and just the concepts of archaeology are something that I've always found interesting on my own as just sort of a layperson to the field. So it's definitely great to see new techniques for trying to bring more people into it and get them starting to think about these different problems that are useful to explore for humanity as a species, both for exploring
[00:36:55] Unknown:
our cultural pasts, but also our cultural present. Right. And and 1 of the neat things of of some of Andrew's work is that he's been looking at a lot of procedural content generation and asking what does it mean to have machine made culture. And, so he's he's done things like lead expeditions into No Man's Sky or he he's, he's been doing code exegesis on the colossal cave adventure, which of course is 1 of the, if not the first, 1 of the first of these immersive worlds, just a text based mediated 1. So there's it's, ArQule Gaming is only gonna grow. I think it's going to become a a really neat, vector that unites, you know, humanities research, technological research, software, game design, game the game industry more generally. It'll be exciting to see how it how it plays
[00:37:51] Unknown:
out. And another interesting area of ways that digital and computational capacity are being leveraged in the field of archaeology is in the space of doing 3 d scans of existing structures as a means of trying to preserve them against areas of conflict or, environmental hazards, such as acid rain or earthquakes, things like that. So I don't know if you have any experience of dealing with any of that types of information or, just that general field of archaeological preservation?
[00:38:29] Unknown:
Yeah. So 3 d modeling and photogrammetry and and and model building that way is it's important work. Right? Because you we're getting you're getting better and better resolution. And if you've got laser scanners, that's the the best resolution going. And it can help guide things like restoration work. We, there's a a gentleman here at Carlton, Steve Fye, who's using laser scanning on, a lot of the architectural elements of the houses of parliament for their for their restoration work and, because, you know, years of acid rain and and so on has worn them right away. The we don't talk about photographs as being preservation.
And so I don't know that we should necessarily talk about digital 3 d models as being a kind of preservation because, you know, so much depends on the the And then you get things like the, the palm Palmyrene Arch episode where the so this in in Palmyra, there's this triumphal arch that was blown up by ISIS, and it had been turned into a 3 d model a few years earlier, but that 3 d model certainly didn't have the resolution or the detail that is capable now. And it was used to mill blocks of stone to create a, you know, a 1 third or 2 third scale model, and everybody said, yay. We've got we've preserved this, but the the model had a lot less of the the the detail.
And then we then people were putting this model up in places like Trafalgar Square to say, yay. We've we've preserved this and and take that Isis. But here it sits in the middle of empire. Right? Trafalgar Square is the center and heart of British Empire, and it's at diminished scale in that context. So looking at it, what kind of message are people meant to to understand here? Right? It's certainly not for the Syrians, and it really kind of, you know, takes that Western colonial gaze again and and brings it, really illustrates how a lot of our digital technologies do reproduce a lot of the ways of looking in the west that are so problematic. I mean, the the arch was then taken down and was displayed in Washington in the center of the mall, I think it was. So digital preservation, taking these 3 d models, I think can be really useful and powerful when there's a really good research case behind it.
And sometimes, you know, sometimes these 3 d models might be all that we can make. When you look at the museum fire in Rio, you can stitch together models from tourist photos if you've got enough of them. Gives us a sense of what has been lost, but it's never it's always going to be lesser. I don't know. It's there's a lot of debate in archaeology about the value of of this kind of work in terms of things like preservation and and so on. I mean, archaeology destroys what it excavates. Right? So the record that exists in our archive is the only representation of this material left. So in a way, archaeology creates its own subject matter, but it also destroys the physical stuff.
Yeah. I'm just thinking out loud and and riffing on on different thoughts, but it's we I think we have to be careful if we equate 3 d model building with preservation. It's it's important work. It can be very useful. But if we we think about it in preservation, then we gotta we get into questions of, well, who does it belong to? Ownership and empower and control, which archaeology has a really dirty backstory in terms of colonialism and and taking and stealing materials that don't necessarily belong to us, western white folk. So, you know, archaeology is always about the present as much as it is about the past.
[00:43:08] Unknown:
Yeah. I really appreciate that context and perspective as somebody who is not as continuously steeped your point about ethics also brings an interesting parallel in terms of the discussions that are happening in the areas of software engineering and data science of trying to determine what are the ethical responsibilities of practitioners in those fields in terms of particularly with data, trying to account for bias that's inherent to the data and the ways that the algorithms are applied, or software engineering, the ways that the interfaces are created and the ways that the, information is used within these systems and the ways that it can support or subvert or create various power structures and create disparities in opportunity or equality. So, this is Exactly. Yeah.
[00:44:08] Unknown:
Yeah. The first question has to always be, who can this hurt? Right? Because that's that's the fundamental function of the computer is to copy, is to replicate. Right? So the decisions that we take on their own, 1 decision, 2 decisions, might not have too much impact. But then when we build that into our systems and the systems replicate and they they intersect with other systems, the the unintended consequences or the emerging consequences can be very dangerous indeed. So absolutely. We, when we sit down and especially when we're dealing with human with whatever data. Right? You gotta the first question has to be, who can this hurt? How can this hurt somebody?
And if you can't answer that question, then you gotta go ask somebody else because it could. It always can.
[00:45:01] Unknown:
And with that, I'm curious, what are your goals and aspirations for the Odate project as you continue to use
[00:45:11] Unknown:
and improve and evolve it? Well, I hope it doesn't hurt anybody. But my my hope is that people will take it to pieces and find the bits that are useful for them in their context, in their teaching, and that it it smooths the way for more people to be engaged with this stuff and and and brings more eyes to to these different problems. Because if if we can increase people's literacy about how these, you know, computational approaches and the the consequences, you know, as a society, that's all to the good and and it's critical and I think we need that. But in terms of archaeology and history, I think we would be doing better history, better archaeology too.
So, yeah, big ambitions, big dreams. I want to improve everything for everyone always, but I would be pleased if people started using it in their undergraduate 1st year courses.
[00:46:13] Unknown:
I also think that it's a useful resource, particularly given the openness of it for anybody who is curious in exploring what archaeology can provide in terms of cultural understanding and knowledge and just ways for people to experiment and engage with the field even just as a layperson?
[00:46:35] Unknown:
Well, absolutely. I I'm I I also teach in the the public history program here, and so I kind of regard my teaching as a as an extension of that as well. The the teaching that we do has to spill out beyond our walls if we want people to actually understand and value what goes on in our work.
[00:47:00] Unknown:
Are there any other aspects of digital archaeology or traditional archaeology or your work with the Odate project that we didn't cover yet which you think we should discuss before we close out the show?
[00:47:12] Unknown:
I I think we've covered a lot of ground, a lot of territory. I I enjoy doing this. I get a huge kick out of people reusing my stuff and and when it triggers new ideas or or it opens up doors for people. So I, you know, I welcome your listeners to take a look at it and and, and join us in pushing it forward.
[00:47:42] Unknown:
And for anybody who wants to follow-up with you or get in touch directly, I'll have you add your preferred contact information to the show notes. And so with that, I'll move us on into the picks. And this week, I'm going to pick a set of headphones that I got recently. They're a pair of noise cancelling earbuds that were fairly inexpensive, and I commute in on the train every day. So it's nice to be able to have that additional level of background noise reduction as I'm just sort of making my way to and from work. So they're, they're by a company called Dautronics.
I'll add the link to the show notes. And so with that, I'll pass it to you, Sean. Do you have any picks this week?
[00:48:23] Unknown:
I've just started reading the latest Ian Rankin novels. So, if you don't know Ian Rankin's work it's detective novels police procedurals featuring a Scottish policeman named John Rebus mostly set in Edinburgh. Great stuff, atmospheric stuff. Just, can't put it down.
[00:48:46] Unknown:
Alright. Well, thank you for taking the time today to join me and discuss your work and experience with the Odate and with digital archaeology. It's been an interesting discussion, and I hope that it opens some doors for people as you mentioned. So thank you for that, and I hope you enjoy the rest of day. Oh, thank you for the opportunity. I've enjoyed it.
Today, I'm interviewing Sean Graham about his work on the Eimeline Digital Archaeology textbook. Hello, and welcome to podcast dot in it, the podcast about Python and the people who make it great. When you're ready to launch your next app or want to try a project you hear about on the show, you'll need somewhere to deploy it. So say hi to our friends over at Linode. With 200 gigabit private networking, scalable shared block storage, node balancers, and a 40 gigabit public network, all controlled by a brand new API, you've got everything you need to scale up. Go to pythonpodcast.com/linode, l I n o d e, to get a $20 credit and launch a new server in under a minute. And don't forget to thank them for their continued support of this show.
And if you're like me, then you need a simple and easy to use tool to keep track of all of your projects. Some project management platforms are too flexible, leading to confusion of workflows and days' worth of setup, and others are so minimal that they aren't worth the cost of admission. After using Clubhouse for a few days, I was impressed by the intuitive flow. Going from adding the various projects that I work on to defining the high level epics that I need to stay on top of and creating the various tasks that need to happen only took a few minutes. I was also pleased by the presence of subtasks, seamless navigation, and the ability to create issue and bug templates to ensure that you never miss capturing essential details.
Listeners of this show will get a full 2 months for free on any plan when you sign up at python podcast.com/clubhouse. So help support the show and help yourself get organized today. And don't forget to visit the site at pythonpodcast.com to subscribe to the show, sign up for the mailing list, and read the show notes. And don't forget to keep the conversation going at python podcast.com/chat. Registration for PyCon US, the largest annual gathering across the community, is open now. So don't forget to get your ticket, and I'll see you there. Your host as usual is Tobias Macy. And today, I'm interviewing Sean Graham about his work on the online digital archaeology text book. So, Sean, could you start by introducing yourself?
[00:02:10] Unknown:
Hi. I'm Sean Graham. I'm an associate professor of digital humanities in the history department at Carleton University in Ottawa, Canada.
[00:02:18] Unknown:
And do you remember how you first got introduced to Python?
[00:02:21] Unknown:
I I think what happened was when I first started at Carlton, I my my computational background was in simulation modeling in agent based modeling. And I I'm not particularly good at any of this stuff. And I was using a platform called Metlogo, which was originally developed to teach high school students about complex systems and complexity theory and and that sort of thing. And when I joined the history department as an archaeologist, I felt like a bit of an imposter, and I needed to quickly learn everything that was going on in other facets of history and archaeology and to figure out how I could fit into everything that was going on here. And a little bit of googling around later, I came across version 1 of the programming historian, which was an online textbook in series of tutorials by Bill Turkel at Western at the University of Western Ontario and Adam Krimbill.
And he had a sequence of common tasks that a historian might want to do programmatically, and they were all written in Python. So I sat down and worked my way through those, and it's been, you know, how fast can I search for something on Stack Overflow? But it was very much courtesy of of Bill and Adam's original work, which came out in about 2, 008, I think. So I started here in 2010.
[00:04:01] Unknown:
Well, it's, opportune happenstance that Python has continued to grow in popularity and use, which I'm sure has simplified some of your ongoing efforts to apply programming practices to your work in the archaeology department and or, sorry, in the history department as an archaeologist?
[00:04:20] Unknown:
Oh, yeah. Very much. And it's it's interesting because everything you know, there are fashions in everything, and there are fashions in history as well as in archaeology. And there there's been a lot of interest in in mapping and web mapping and historical GIS. And things like ArcGIS have rather expensive licenses. So open GIS software, things like QGIS, which has a lot of modularity built into it and and can be scripted in Python and so on, means that trying to stay 1 step ahead of where my students are interested keeps pushing me to to learn new things.
[00:05:04] Unknown:
Yeah. I've had occasion to use QGIS myself for some physical installations of IoT sensors in a past job, and it was definitely useful for being able to have a open and accessible platform to tinker with and build up maps of where the different sensors were and try to create overlays of what the, network paths were.
[00:05:27] Unknown:
Absolutely. And the the intersection of Python with all sorts of different open source platforms has been a real boom for me because I can't I'm not allowed to, assign a particular software package as a required purchase to students. I can I can put down, you know, purchase this $200 textbook without issue? But if I say you you have to buy a $20 subscription to something, that I'm not allowed to do. So we we need to do some catching up. But in the meantime, it pushes me to use a lot of open source stuff, which is great because then the students learn how to fend for themselves and figure out how to find the packages or the software or the platforms or whatever that can help them reach their research goals.
[00:06:17] Unknown:
1 of the few cases where the unintended side effects of bureaucracy ends up being positive. Absolutely. Yeah. And so as part of your work as a professor in archeology and in the history department, you have collaborated on this online digital archaeology textbook or Odate for short. So can you start by giving a bit of an explanation of at least how you see the definition of digital archaeology?
[00:06:47] Unknown:
Sure. For me, digital archaeology is not just about using computational power to ask archeological questions or to answer archeological questions. It also involves thinking through what these tools and approaches do to how I understand or conceive the past as well. So in the Open Digital Archaeology textbook, we have a lot of support, I suppose, for using digital tools to deform our archaeological material or deform our perspective on what we're looking at, to think through creative uses, so computational creativity as well as straightforward, you know, spatial statistics or or what have you. So it sits in a broader digital humanities framework.
Archaeology has been using things like GIS for 40 years and there's any amount of support for teaching GIS, but there's much less support out there for trying to say you're interested in how visitor impacts on an archaeological site in a national park and maybe what you wanna be doing is scraping Instagram and studying comments for what people are attending to and what what they find is important or or meaningful. So that's the kind of thing you can also learn courtesy of of our textbook. And in teaching or trying to teach these kinds of methods in my own classes, I end up wasting 2, 3 weeks on tech support and getting everybody on the same page.
I have to rely very much on my students bringing their own device, but I can't be supporting 3 different flavors of Windows and machines that might have 2 gigs of RAM and or 7 gigs or or the odd person with a Linux box or, you know, Mac machines from 4 or 5 years of of of different specs. So with Odate, we also the the original vision was to have an entire virtual machine that could be installed with 1 click, and the textbook would be inside the virtual machine as well. That proved to be really difficult because of the security implications and the university not wanting to open ports to to anybody to be running code on our servers.
I was developing all of this and proposing all of this right in the middle of a ransomware attack at our university. So, you can imagine how nervous what I was proposing was was making people. But it was around about that time that Tim Sherritt, who is a digital historian in Australia, sent me a direct message on Twitter saying, have you seen this binder service? And that that really enabled everything to take right off because I could just package up targeted exercises or packages or or, Jupyter Notebooks in a single repo, 1 click, boom, binder, virtual environment, and we were off to the races.
So it's, it it was very much different what I ended up building than what I initially proposed. And this was funded by an outfit called Ecampus Ontario, which is a an agency of the Ontario Provincial Government that wanted to, wants to support Open Educational Resources. And in their initial, description of the call for proposals it was clear that they imagined digital textbooks as being basically PDFs that a person could download and and maybe print out and put into a course pack. So an online digital textbook with Jupyter notebooks and binder powered environments, really pushed the envelope for what they were imagining.
[00:11:07] Unknown:
And in the end result, it ends up being much easier to keep things relevant and up to date as well because you don't have this static asset of a PDF, for instance, that somebody might download and can quickly have issues with various different versions floating around or not knowing how up to date something is, or if new versions of packages come out and you're having source code embedded within the PDF that might not actually be accurate anymore versus this binder environment that is hosted, that is always going to be consistent for everybody who accesses it within a given time frame. So I imagine that has greatly simplified your process of teaching these systems as well. Oh, yeah. Absolutely.
[00:11:54] Unknown:
And and it's great too because, you know, when we revealed this to the world, I said, and I identified Twitter and via my blog, you know, I said, look, this is there's gonna be rough edges. There's gonna be things that don't work. There's gonna be things that get dated or packages are gonna change. That's okay. That's part of the learning. That's part of what we want you all to understand. Here's how you can contribute. Here's how you can suggest changes. Here's how you can tell us when things are broken, this is how we can fix it. And so it becomes becomes very much a living document. And we've already had some poll requests from people external to the project improving, our section on databases, for instance, adding adding sections on graph databases.
But even 1 of my even 1 of the Jupyter Notebooks that we built for our initial launch, it was, it was using or it was trying to teach about principles of linked open data and how to do sparkle queries and so on. On. And we were using the endpoint from the British Museum. The British Museum totally pulled the plug on all of that the people who'd been supporting it and and championing it at, at their institution. And so that endpoint went down and is no longer supported. So, you know, things change quickly in in the the world of museums and and, cultural institutions as far as their their online stuff goes.
So if if we were building things in a static sense, that would be an enormous disaster for us. But because we anticipate, we expect this kind of stuff to happen in the first place, the totally, the repos and notebooks and binders, turns it into a a learning experience for everybody. So it's it's authentic. It's real.
[00:13:45] Unknown:
And 1 thing that I found interesting in the process of reading about the textbook and some of your blog posts on the matter and looking through the material is the concept of a perpetual beta that you are supporting for this work, which means that there as you said, you decided to go ahead and launch it, even though it wasn't necessarily perfect in your mind, but recognizing that it will never reach that state of perfection, especially if you never release it to the world. So wondering if you can talk a bit about your philosophies of how you view the textbook as this living entity and some of your hopes for how the community will engage with it both in terms of the software community and the broader academic community?
[00:14:33] Unknown:
Yeah. Well, the part of the advantage of a perpetual beta framework is I'm I'm not expert in every single thing that we've covered in the in the volume, and I never could be. I have I have deep experience in a extremely narrow aspect of of all of this. So I I don't think we could ever reach the kind of perfection that you would want for or that we have been trained to expect for print publications. Right? You you put something out in prints and you never revisit it. By by focusing it or branding it, maybe that's a better word, but by saying to people, look, it's perpetual beta means that my hope is that when people encounter things that are a bit wonky or not as clearly explained as we might want, that they'll take it in the spirit in which it is offered and try to to improve things.
And in that sense, if people fork it and it ends up going in in many different directions, that's that's good. I I don't wanna be prescriptive about what digital archaeology is or could be, but rather the people who who need it or who want it and take it forward become in the driver's seat. And I know that it's it's gonna mean that I have to always keep an eye on what's going on, and, eventually, maybe I'll be able to handle the the official reins to somebody else. But it also opens up teaching opportunities for me too with my my graduate students, my undergraduates, about what it means to to build community or to to try and keep something like this as current as possible.
So Perpetual Beta, in 1 sense it's a bit me covering my ass and recognizing that I couldn't hope to put out something that covers all the bases and is perfect the first time. But I, more seriously, I I regard it as an open invitation, and I hope that people will, you know, will accept it in the in the spirit in which it is offered.
[00:16:52] Unknown:
And in terms of the subject matter that you cover within the textbooks and the notebooks that go with it, I'm curious what you have found to be some of the most critical lessons for your students in order to help them be effective and successful archaeologists?
[00:17:09] Unknown:
The, the the idea about failing productively and and how to to recover from things that don't work or things that collapse on you. My my students coming at me in the history department and to a lesser degree in archaeology, they they have been trained for years to just accept what the computer gives them and to open the hood and try and get it to do something else to, you know, the kind of things that we, you know, growing up in the eighties, you sit down at a Commodore 64 and you turn it on and it says ready and it just blinks and it does nothing unless you start telling it what to do. The my students, at the very least, come with a huge degree of anxiety about about trying anything other than what the software already gives them. So they very much live inside the boxes that are are provided.
Teaching people that it is okay to open the hood, that it's okay to take control of of your machine and to push back against the decisions that other people have made in their code so that you can do your research more honestly or more effectively or in new directions. That's that's a really hard thing to teach. So in the in the textbook, we put a lot of effort into talking about a productive fail mindsets, about open and reproducible code, about sharing your code for replicability and reproducibility. These are things that we don't normally teach in at least on the humanity side of the, of the quad.
So the tech will always be changing. But if you've got that good mindset from the outset, you'll be able to figure out what it is you need out of the tech and where to find answers and how to roll with it. And that, I think, is probably the single most useful thing that comes out of this whole experience.
[00:19:22] Unknown:
In terms of their experience once they graduate and get involved in actually doing archaeological research or field work, I'm wondering what you view as being the level of necessity of that computational comfort and experience as they progress throughout their careers?
[00:19:41] Unknown:
I think it's only gotta be to the good. If we can teach people how to figure out what they need to know, when they need to know it, and where to go in order to get the solutions. I mean, I spend huge amounts of time on Stack Overflow. Thank you to those folks. Then just getting in the habits of thought and understanding how code works and how to read it and and what to where to troubleshoot. Like, those are the things that, to my mind, are are more valuable than experience with any particular package at a time. Right? Like, it's it's the problem of of GUI based software packages in the first place. If the the GUI changes, then you don't know what to do anymore.
So does does that answer that question?
[00:20:33] Unknown:
It does in part, but I was also curious in the broader sense of the amount of crossover of the computational tools that are required for doing archaeological research and field exploration for people who are engaged in that in a professional context both now and as you see it progressing into the future?
[00:20:56] Unknown:
Crossover from what we're teaching to their professional lives?
[00:21:00] Unknown:
Just in terms of archaeological research, I'm just curious, like, how prevalent using software packages and computational techniques is applied versus spending time digging up artifacts or researching physical structures, things like that. Right. Right.
[00:21:19] Unknown:
Well, in in commercial archaeology right now and I and I mean I have to be honest, I haven't worked in commercial archaeology in quite some time. I actually have very little experience in commercial work but there's computational work happening all the time Databases, mapping, GIS, public outreach for all of it. I'd say the the computational work is more more the bread and butter than the the actual field work as it were because as the field work happens, the the material is being digitized as the field work goes on. Right? Things are being excavated, things get recorded into various databases, the the whole analytical side that happens the field work is you know maybe 1 third of what goes on and it's all the behind the scenes stuff, the, the assessments that happen before field work goes on, the predictive work, the the archival work to, work out the probabilities of what kinds of materials you might reasonably expect. Like, that sort of work is very computational and probably occupies far more of the the working life of of people in commercial work than than than field work does.
[00:22:47] Unknown:
And in terms of the data sources and data formats that would commonly be encountered, you mentioned some of these geographical surveys and GIS information, digitization of some of the artifacts that might be recovered, whatever formats that might take, and then also recording of metadata. And, I'm sure there are time aspects to when the artifacts are recorded or when the observations are taken. So I'm just wondering if you can talk through those different types of data and the sources and locations that they might be found for people who are engaged in performing these computational analysis of all of that information.
[00:23:29] Unknown:
Right. Well, it's, it's it's a god awful mess. There is material trapped in image based PDFs. There's stuff in old versions of Excel. There are XML, JSON, CSV, FileMaker, God help us all, Word documents. There there's there's no there there are archaeological repositories for digital work, and it's not necessary not necessarily true that every academic archaeologist or every commercial archaeologist is putting materials into these repositories. Some of the bigger ones are open context, open context dot org, which tends to have everything represented in JSON. Then there's TIDAR, which is out of Arizona State University, I believe, and they are more of a they they will take your documentation in all sorts of different formats, whereas OpenContacts works with you as a data publisher to get your data into a format that can be exposed through their API and can be recombined and reused with other stuff. There's the archaeological data service in the UK, which again will take materials in lots of different formats, everything from field notes to to tables.
So you you have you know, 1 of the things that we, 1 of the binders that we have in Ode uses an R package called Tabulizer to, to try and, when you encounter data locked in a PDF to to get it into a data frame that you can then use. So there's there's some folks are using graph databases. Some people are using, all sorts of different schemas. It's, you you can you can encounter data in all sorts of different formats. A friend of mine who's, I forget what the the proper title is, but it is states a state archaeologist responsible for a lot of the data management of things. She encounters material kept in word files, material kept in old shoe boxes. It's it's a wild west of data.
[00:25:59] Unknown:
And in terms of the notebooks that, you and your collaborators have created to accompany the textbook, there's a fairly even mix year of R and Python code. So I'm wondering what you have found to be your personal guidelines for when to use each language to its greatest benefit.
[00:26:20] Unknown:
I'm not sure that I have anything really useful to say on that because I'm I'm a bit of a magpie when it comes to to code. If if I can get it to work and it does roughly what it is I need it to do, then that ends up being the the language or approach that I I end up using. So I got into using R via text analysis and topic modeling and that sort of thing. A lot of archaeologists are more comfortable with Python than they are with R. That said, Ben Marwick at, Washington State University has a huge compilation of R packages that are useful for archaeology. Matt Harris, is another 1 who's very, very handy with our can do all sorts of interesting things with it.
So, really, if it does the job you needed to do, then that's the package you use. Yeah. Not the most helpful approach, I suppose. But as long as your code is is well commented out and you understand what it's doing, then either or, I suppose.
[00:27:33] Unknown:
Yeah. And I think that just in the broader sense of software, once you have enough of a foundational understanding of what's happening, then the syntax to some degree starts to melt away, and it's much better to be pragmatic than dogmatic in your approach to when to use which tool.
[00:27:52] Unknown:
Absolutely. Absolutely. I'm in I'm entirely self taught. I, I screw up and make errors all the time. So I really appreciate people who put out packages that are are cleanly explained and and commented and have examples that I can work through and think, okay. I see it. This is how their data is organized. I can get my data organized like that. I can use their sample code. Now what can I plug this in with to get to the next stage that I need it to be at? So the the IR OpenSci folks, they they're they've got lots of really great packages, and and I really appreciate what they're up to as well.
[00:28:34] Unknown:
Yeah. And particularly in recent years, there has been a lot more work done to ease the interoperability between R and Python in the data analytics space with things like the arrow project to make data frames interoperable between R and Pandas, and just a a general approach within the community of, academics and scientists and data scientists to lower the friction and just make collaboration easier regardless of the language community?
[00:29:07] Unknown:
Absolutely. And I'm I haven't I haven't explored that project very much yet, but that's that kind of work is is really important, but it still requires an interpreter to get it down to the level of where so many of our students are. So that's something I'm gonna have to put on the agenda for future updates to, to Odate.
[00:29:30] Unknown:
And in terms of your overall experience and career, I'm curious how the skills and tools and thought models of software engineering have influenced your views and approaches to the way that you conduct research and education in in the realm of archaeology and history?
[00:29:51] Unknown:
Well, I I feel in some senses, I feel like I'm 1 step removed because I'm I'm very much influenced by people like Carl Boetticher, and I'm not entirely sure I pronounced that correctly, who was big on open notebook science and and and Ben Marwick, who's an archaeologist, or Caleb McDaniel, who's a historian of these these ideas of replicable code chunks and and being open about our processes entirely as academics. So, the degree to which the actual software community influences. I'm sure it does. I I'm sure it is because I'm, you know, I'm picking over carefully through people's blogs and Stack Overflow posts and trying to figure out my own immediate concerns.
And I expect I I'm picking it up through osmosis, but I'm not I don't know that I could point to any single individual or or group of of people.
[00:31:00] Unknown:
And in terms of the ways that, software and analysis and computational capacity are applied to the field of archaeology. What have you found to be some of the most novel or engaging ways to make use of that?
[00:31:20] Unknown:
Well, I'm really interested in neural networks, for computer vision at the moment, and I I'm fascinated by things like generative adversarial networks and people exploring the the kind of latent space of images. And I'm I I have a research project on at the moment that looks at people who buy and sell human remains and trade them on, via social media and using computer vision to try and work out some of the rhetorical tropes of the visual composition of these images. Like, what is it that people are how are they signaling that things are for sale or what kinds of ideas that have you know, 1 1 thing that we're starting to notice is how when people are trying to signal something's for sale, but not actually stayed as much in terms of the comments, the the images end up being composed rather like how we would imagine a museum would display human remains of the, like, the 19 thirties or forties.
So sort of an Indiana Jones visual motif, coming through in the images, which, you know, you might pick up on if you were staring at pictures all day. But when we do it at scale with convolutional neural network, it it stands out an an awful lot more. So that kind of thing is, you know, where I'm at at the moment. Other people are using them to pick out features in, field photographs. So as you're excavating you're taking photos of each level or feature or context as you dig down. So every event that happens in the formation of a site is a context and you can peel these back like layers of onion or and the way that they intersect is a sort of three-dimensional, 4 dimensional puzzle.
And there are some folks who are feeding 100 or 1000 of these kinds of photographs into neural networks to see what, what can we pick up when we let the computer do the looking rather than us. So, you know, interesting things like that.
[00:33:47] Unknown:
And another piece that I saw referenced as I was preparing for this show is the concept of archeogaming. So I'm wondering if you can talk a bit about what that is and some of the ways that it's applied.
[00:34:00] Unknown:
Right. So so archeogaming as a term was coined by Andrew Reinhart, who is a classical archaeologist and, archaeological publisher. And he was also involved with Bill Caraher and a few others in the Alamogordo excavations of the Atari burial ground. So if you've seen the game over documentary on Netflix about those excavations, you will have seen Andrew Reinhart. And he's actually doing a PhD in this at the moment, which is quite exciting. And he and I and a number of others for a couple of years have been exploring ideas of what does it mean for humans when so much of our culture these days is digitally mediated or takes place within digital spaces.
And archaeology is about the material expression of human culture whether it's in the remote past or the current current day. So if we do archaeology inside these digital spaces, what does that look like? How do we translate the methods and perspectives of dealing with material culture into, into digital culture? So archaeogaming can be the application of archaeological perspectives within constructed digital spaces. It can be looking at the representation of human past or material culture through video games. So 1 aspect of archeo gaming also intersects with archeological ethics when you think about loot looting mechanics and loot boxes and, archaeological tropes within video games.
There's a there's another group at the University of Leiden who have been doing a lot of work in in ArQube Gaming and they run an annual conference called the, the interactive pasts conference. So it's, it's kind of an exciting field. Right? Because it's a new new avenue for for human, for archaeology to try and understand the human condition. Yeah. Humans have been making virtual worlds since the the Neolithic. Right? You look at the caves of Lascaux, you look at, you know, all the way up to Disneyland, that's another virtual world. So now it's moving into these, crafted virtual worlds that we interact with through the screen.
[00:36:25] Unknown:
Yeah. It's definitely a very fascinating area and just the concepts of archaeology are something that I've always found interesting on my own as just sort of a layperson to the field. So it's definitely great to see new techniques for trying to bring more people into it and get them starting to think about these different problems that are useful to explore for humanity as a species, both for exploring
[00:36:55] Unknown:
our cultural pasts, but also our cultural present. Right. And and 1 of the neat things of of some of Andrew's work is that he's been looking at a lot of procedural content generation and asking what does it mean to have machine made culture. And, so he's he's done things like lead expeditions into No Man's Sky or he he's, he's been doing code exegesis on the colossal cave adventure, which of course is 1 of the, if not the first, 1 of the first of these immersive worlds, just a text based mediated 1. So there's it's, ArQule Gaming is only gonna grow. I think it's going to become a a really neat, vector that unites, you know, humanities research, technological research, software, game design, game the game industry more generally. It'll be exciting to see how it how it plays
[00:37:51] Unknown:
out. And another interesting area of ways that digital and computational capacity are being leveraged in the field of archaeology is in the space of doing 3 d scans of existing structures as a means of trying to preserve them against areas of conflict or, environmental hazards, such as acid rain or earthquakes, things like that. So I don't know if you have any experience of dealing with any of that types of information or, just that general field of archaeological preservation?
[00:38:29] Unknown:
Yeah. So 3 d modeling and photogrammetry and and and model building that way is it's important work. Right? Because you we're getting you're getting better and better resolution. And if you've got laser scanners, that's the the best resolution going. And it can help guide things like restoration work. We, there's a a gentleman here at Carlton, Steve Fye, who's using laser scanning on, a lot of the architectural elements of the houses of parliament for their for their restoration work and, because, you know, years of acid rain and and so on has worn them right away. The we don't talk about photographs as being preservation.
And so I don't know that we should necessarily talk about digital 3 d models as being a kind of preservation because, you know, so much depends on the the And then you get things like the, the palm Palmyrene Arch episode where the so this in in Palmyra, there's this triumphal arch that was blown up by ISIS, and it had been turned into a 3 d model a few years earlier, but that 3 d model certainly didn't have the resolution or the detail that is capable now. And it was used to mill blocks of stone to create a, you know, a 1 third or 2 third scale model, and everybody said, yay. We've got we've preserved this, but the the model had a lot less of the the the detail.
And then we then people were putting this model up in places like Trafalgar Square to say, yay. We've we've preserved this and and take that Isis. But here it sits in the middle of empire. Right? Trafalgar Square is the center and heart of British Empire, and it's at diminished scale in that context. So looking at it, what kind of message are people meant to to understand here? Right? It's certainly not for the Syrians, and it really kind of, you know, takes that Western colonial gaze again and and brings it, really illustrates how a lot of our digital technologies do reproduce a lot of the ways of looking in the west that are so problematic. I mean, the the arch was then taken down and was displayed in Washington in the center of the mall, I think it was. So digital preservation, taking these 3 d models, I think can be really useful and powerful when there's a really good research case behind it.
And sometimes, you know, sometimes these 3 d models might be all that we can make. When you look at the museum fire in Rio, you can stitch together models from tourist photos if you've got enough of them. Gives us a sense of what has been lost, but it's never it's always going to be lesser. I don't know. It's there's a lot of debate in archaeology about the value of of this kind of work in terms of things like preservation and and so on. I mean, archaeology destroys what it excavates. Right? So the record that exists in our archive is the only representation of this material left. So in a way, archaeology creates its own subject matter, but it also destroys the physical stuff.
Yeah. I'm just thinking out loud and and riffing on on different thoughts, but it's we I think we have to be careful if we equate 3 d model building with preservation. It's it's important work. It can be very useful. But if we we think about it in preservation, then we gotta we get into questions of, well, who does it belong to? Ownership and empower and control, which archaeology has a really dirty backstory in terms of colonialism and and taking and stealing materials that don't necessarily belong to us, western white folk. So, you know, archaeology is always about the present as much as it is about the past.
[00:43:08] Unknown:
Yeah. I really appreciate that context and perspective as somebody who is not as continuously steeped your point about ethics also brings an interesting parallel in terms of the discussions that are happening in the areas of software engineering and data science of trying to determine what are the ethical responsibilities of practitioners in those fields in terms of particularly with data, trying to account for bias that's inherent to the data and the ways that the algorithms are applied, or software engineering, the ways that the interfaces are created and the ways that the, information is used within these systems and the ways that it can support or subvert or create various power structures and create disparities in opportunity or equality. So, this is Exactly. Yeah.
[00:44:08] Unknown:
Yeah. The first question has to always be, who can this hurt? Right? Because that's that's the fundamental function of the computer is to copy, is to replicate. Right? So the decisions that we take on their own, 1 decision, 2 decisions, might not have too much impact. But then when we build that into our systems and the systems replicate and they they intersect with other systems, the the unintended consequences or the emerging consequences can be very dangerous indeed. So absolutely. We, when we sit down and especially when we're dealing with human with whatever data. Right? You gotta the first question has to be, who can this hurt? How can this hurt somebody?
And if you can't answer that question, then you gotta go ask somebody else because it could. It always can.
[00:45:01] Unknown:
And with that, I'm curious, what are your goals and aspirations for the Odate project as you continue to use
[00:45:11] Unknown:
and improve and evolve it? Well, I hope it doesn't hurt anybody. But my my hope is that people will take it to pieces and find the bits that are useful for them in their context, in their teaching, and that it it smooths the way for more people to be engaged with this stuff and and and brings more eyes to to these different problems. Because if if we can increase people's literacy about how these, you know, computational approaches and the the consequences, you know, as a society, that's all to the good and and it's critical and I think we need that. But in terms of archaeology and history, I think we would be doing better history, better archaeology too.
So, yeah, big ambitions, big dreams. I want to improve everything for everyone always, but I would be pleased if people started using it in their undergraduate 1st year courses.
[00:46:13] Unknown:
I also think that it's a useful resource, particularly given the openness of it for anybody who is curious in exploring what archaeology can provide in terms of cultural understanding and knowledge and just ways for people to experiment and engage with the field even just as a layperson?
[00:46:35] Unknown:
Well, absolutely. I I'm I I also teach in the the public history program here, and so I kind of regard my teaching as a as an extension of that as well. The the teaching that we do has to spill out beyond our walls if we want people to actually understand and value what goes on in our work.
[00:47:00] Unknown:
Are there any other aspects of digital archaeology or traditional archaeology or your work with the Odate project that we didn't cover yet which you think we should discuss before we close out the show?
[00:47:12] Unknown:
I I think we've covered a lot of ground, a lot of territory. I I enjoy doing this. I get a huge kick out of people reusing my stuff and and when it triggers new ideas or or it opens up doors for people. So I, you know, I welcome your listeners to take a look at it and and, and join us in pushing it forward.
[00:47:42] Unknown:
And for anybody who wants to follow-up with you or get in touch directly, I'll have you add your preferred contact information to the show notes. And so with that, I'll move us on into the picks. And this week, I'm going to pick a set of headphones that I got recently. They're a pair of noise cancelling earbuds that were fairly inexpensive, and I commute in on the train every day. So it's nice to be able to have that additional level of background noise reduction as I'm just sort of making my way to and from work. So they're, they're by a company called Dautronics.
I'll add the link to the show notes. And so with that, I'll pass it to you, Sean. Do you have any picks this week?
[00:48:23] Unknown:
I've just started reading the latest Ian Rankin novels. So, if you don't know Ian Rankin's work it's detective novels police procedurals featuring a Scottish policeman named John Rebus mostly set in Edinburgh. Great stuff, atmospheric stuff. Just, can't put it down.
[00:48:46] Unknown:
Alright. Well, thank you for taking the time today to join me and discuss your work and experience with the Odate and with digital archaeology. It's been an interesting discussion, and I hope that it opens some doors for people as you mentioned. So thank you for that, and I hope you enjoy the rest of day. Oh, thank you for the opportunity. I've enjoyed it.
Introduction to Sean Graham and Digital Archaeology
Sean Graham's Background and Introduction to Python
Defining Digital Archaeology
Challenges and Solutions in Teaching Digital Archaeology
Philosophy of Perpetual Beta in Digital Textbooks
Critical Lessons for Archaeology Students
Computational Tools in Archaeological Research
Data Sources and Formats in Archaeology
Choosing Between R and Python
Impact of Software Engineering on Archaeology
Novel Applications of Computational Tools in Archaeology
Understanding Archeogaming
3D Scanning and Preservation in Archaeology
Goals and Aspirations for the Odate Project
Public Engagement and Open Resources
