10 minute read

Ever since I created this website, I have gotten questions about what the term “Computational Archaeology” actually means. Most people seem to generally understand that it has something to do with using computers and also something to do with archaeology. But, beyond that, ideas seem to diverge fairly rapidly.

Computational archaeology?
A computer hard at work on something archaeological. Is this all there is to "Computational Archaeology?"

The fact is that when I created this site, I indeed had a very specific understanding of what “Computational Archaeology” meant to me. Back then, I thought that I had, in fact, coined the term. In the time since then, however, I have not only come to discover that I did not invent the term, but that it has been used extensively by others, and in some contexts that differed from my originally intended use on this site (albeit certainly with overlap). I will get to my own thoughts a little later in this post, but let’s first start, as everything does these days, with a Google search.

Definitions from around the web:

As of the time of this blog post, my website sits at the number 5 position in the Google search ranking for “Computational Archaeology.” First in the list is the wikipedia post on Computational Archaeology, then Kristina Georgieva’s Wordpress blog on becoming a Computational Archaeologist, a “How Stuff Works” article, and a link to the wonderful MSc Program in Computational Archaeology at UCL. Looking across these top five Google hits, other places around the web, conversations and paper titles at professional meetings, and in the literature, I have noticed a few trends in the way “Computational Archaeology” is defined:

  1. The use of digital data in archaeological research.
  2. The general use of computers in archaeological research.
  3. The use of computational algorithms to fill gaps in archaeological data in ways that can’t be done with traditional techniques.
  4. Data mining and multidimensional analyses of archaeological data.
  5. The use of simulation models and artificial intelligence to understand the processes and dynamics of past societies.
  6. The use of computer visualizations to create immersive virtual experiences of archaeological sites and materials

Various definitions will emphasize some of those subjects more than others, however, and may even ignore the rest of the list. All definitions emphasize the use of computers, but do so in different ways. I have noticed the following emphases on computer-based tools in various definitions of “Computational Archaeology”:

  1. GIS and geospatial data analysis.
  2. Digital imagery and photogrammetry.
  3. Statistical computing and quantitative analysis.
  4. Data visualization.
  5. Agent Based Modeling and simulation.
  6. 3D modeling, virtual reality, and augmented reality.
  7. Video gaming and digital animation.

In the various definitions you will find, there is a lot of talk about the use of “new” and “innovative” approaches and about how the field is not yet fully developed. For example, the Wikipedia article that comes up first in the Google search list describes it as “an emerging discipline” that is “often inadequately represented in archaeological training and education.” While I agree with this to some extent, I also would point out that none of the trends or tools listed above are particularly “new.”

Early computational archaeology.
A Fortran punch card from the 1960's shows how just how "new and innovative" computation is. (By Arnold Reinhold CC BY-SA 2.5, via Wikimedia Commons)

More truthfully, the technology and many of the basic mathematical and computer concepts techniques have been around for quite a long time (my whole life, for example). It may still be true, however, that the impact of these things on mainstream archaeology has not yet reached a saturation point. That is, it is unclear which, if any, of the things listed above are as yet considered part of the “normal” toolkit of the practicing archaeologist.

The oldest, and perhaps most influential organization that I would associate with Computational Archaeology is the venerable Computer Applications & Quantitative Analysis in Archaeology group, or “CAA” for short. The CAA revolves around an annual conference that rotates through various countries from year to year. The first CAA conference was in 1973 – which means that the organization is neither “new” nor “recent.” While I have somewhat shamefully never personally attended a CAA conference, I hear through the grapevine that they really are a true hub of information exchange about the method, theory, and practice of pretty much everything I’ve listed above in terms of trends and tools. The CAA publishes very useful and interesting conference proceedings volumes, as well as a regular journal. In many ways, the CAA is the organization that seems to align most closely to what Computational Archaeology seems to be about.

CoMSES, the Community for Modeling Social-Ecological Systems, is another organization that falls within our growing understanding of a “Computational Archaeology.” While explicitly interdisciplinary, CoMSES has traditionally held a focus on simulating past societies via ABM techniques (and other modeling formalisms). Several of the CoMSES board members, and many community members are archaeologists. Simulation is a major focus of CoMSES. CoMSES was founded back in 2007 and I have been a member for some time now. Is 11 years old still “new” enough to be called “recent?”

Unarguably “recent,” on the other hand, are two interest groups in the Society for American Archaeology (SAA) that overlap with the idea of computational archaeology. The Digital Data Interest Group (founded in 2015? 2014?) focuses on the use and curation of digital archaeological data. The Open Science Interest Group (founded just last year) focuses on the open-sourcing of archaeological analytical methods and the process of doing archaeology (with a heavy focus on computer-based ways of doing this). I belong to both these groups, and while neither purports to cover the entirety of “Computational Archaeology,” they are both certainly a part of that general idea.

Other terms that are in use:

There are two other related terms that are out there circulating: “Cyber Archaeology,” and “Digital Archaeology.” “Cyber Archaeology” is a term coined by Tom Levy at UCSD for his UC Cyber Archaeology Lab. The focus of this lab is on digital data collection and analysis in archaeology.

“Digital Archaeology” is more widely used, and is certainly currently a more popular term than is either “Cyber” or “Computational” archaeology. The Google search results for this term come back with many results, including several “institutes” and “centers” for Digital Archaeology hosted by various universities, by non-profits, and even by some commercial outfits as well.

In this search list also appears the journal, Frontiers in Digital Humanities: Digital Archaeology (FDHDA). I was/am/have been/will be listed as a review editor for this journal (although my name seems to have gone missing from their website at the moment). The journal is currently in a bit of flux, and its scope is quite frankly unclear even to me at this point. However, looking through both the general Google search results for ‘Digital Archaeology” and the list of articles published in FDHDA, “Digital Archaeology” seems to be understood similarly to Tom Levy’s “Cyber Archaeology.” The focus is mainly on recording archaeological data digitally (3D scans, mobile forms, photogrammetry, etc.). Here, there is a clear and large overlap with the growing trend of “Digital Humanities.”

Ok, so how do I define “Computational Archaeology,” then?!

As I intimated at the top of this post, when I decided to use the term “Computational Archaeology” in the name of this site, I had a very well defined notion of it in my head. Let’s first take a look at what the Merriam Webster dictionary tells about the word “Computation”.

Merriam Webster dictionary definition for computation
Accessed 05/09/2018.

In my view, I reserve “Computational Archaeology” for analyses that involve actual calculation. That is, I prefer definition 1a, above. In this sense, “computational” is taken to mean “computing some sort of result”. This, to me, must be more than simply “using a computer,” as in definition 1b, but very often does involve – or perhaps even requires – the use of computers.

I break this down further into two general types of computation:

First, there are analytical computations, which calculate a mathematical solution from input data. This is the realm of “scientific computing,” using software such as R or Scientific Python, as well as GIS analysis with advanced tools like GRASS. Multidimensional statistics, clustering, data-mining, predictive modeling, cross-correlation, geospatial analytics, imagery analysis, etc. are all analyses that I would put in this subcategory of Computational Archaeology. They are generally quantitative, often heuristic, and sometimes algorithmic approaches to explaining or finding patterns in data. This might be seen as the inductive approach to computational archaeology.

Second, are generative computations, which compute novelty from a set of initial conditions. This is the realm of “simulation,” using modeling software like NetLogo or RePast. Here, the calculations being performed by the computer are generating non-linear behavior in highly algorithmic systems of interacting software agents. “Agent-Based Modeling,” “genetic algorithms,” “cellular automata,” “stock and flow models,” “dynamic equilibrium models,” etc., are all within this subcategory. The point here is to be inspired by archaeological and anthropological theory and data to create dynamic, processual simulation models that produce observable behavior from first principles. This might be seen as the deductive approach to computational archaeology.

Anything that is not “computation” according to definition 1a, above, is in my view “Digital Archaeology.” To me, this is a better label for things that involve computers, but which are not actually computing anything. Digital data collection, manipulation, curation, and interaction all fall into this other category. I prefer the term “Digital” archaeology over “Cyber” archaeology simply because it has caught on more – in my mind the two terms are perfectly interchangeable. This a rich and very exciting field, and one that is currently experiencing a renaissance as the price point of technologies such as UAV’s, LiDAR, GPS, Mobile Tablets, and high-resolution cameras has fallen increasingly into the “consumer” realm over the last few years.

So here’s how I break it all down:

Things that you need computers for in archaeology:

Digital Archaeology Computational Archaeology
Digital data collection Quantitative data analysis
Data storage Pattern recognition
Data manipulation Predictive modeling
Data visualization Simulation modeling

Computers and Archaeology:

At the end of this now rather lengthy blog post, I want to emphasize that I am not trying to be a “splitter” here. I do not want to seem to be encouraging a separation between “Computational” and “Digital” archaeology. Instead, I want to show how these two very-much-related subcomponents of computer-based archaeology articulate. On the one hand, digital archaeologists are gathering and encoding incredible amounts of data in digital formats. They are innovating methods of handling, visualizing, and qualitatively understanding these data. They are deeply embedded in the realm of technology and draw their intellectual inspirations from the digital humanities.

On the other hand, computational archaeologists are expanding our ability to quantitatively interrogate data, and to understand dynamics and process that are unobservable in the archaeological record. We need to be doing both of these things. Gathering fancy digital data for fancy digital data’s sake is not enough, even if we can now get a personal feeling for it through VR or gaming. Simulations and deep statistical techniques are meaningless without the appropriate type and scale of data to input or to validate them. In this way, these two approaches compliment and extend the each other’s utility.

To sum up, way back when I chose the name for this site, I had simulation modeling and geospatial analyses on the brain. My work has mainly focused on the creation of artificial societies in the computer for “counterfactual” simulation experiments. I have specifically looked at the use of GIS as a way to model the spatial component of these simulations, and explored the ways in which coupled human and natural systems can be simulated. I am mainly interested in process and dynamics, so generative modeling has always been the most attractive part of “computational” archaeology for me. And really, that is the reason I started using computers in archaeology way back in the day.

The use of drones, photogrammetry, GPS, mobile tablets, and other digital technology has long been a part of my tool kit, however, even if these weren’t the aspects of my work that I have emphasized over the years. Although I still like to call myself a “Computational Archaeologist,” since that’s still where the bulk of my interests lay, it’s pretty clear that I could just as easily be called a “Digital Archaeologist” as well. Moreover, I would be very surprised if there was anyone who did only “computational” archaeology and no “digital” archaeology, or vice versa. You really can’t have one without the other.

Hmmm…. Maybe I should just start calling myself a “Computer-Using Archaeologist?”

This post was inspired from the seminar discussions in my GIS and Imagery Analysis course over the past two years, as well as a “special topics” course on the subject of “Computational Archaeology.”

Leave a comment