Archaeogaming

On January 28, 2018, I posted a full write-up on what I consider to be landscape archaeology of a valley and village in Skyrim, which sought to apply both traditional and new methods of experiencing a designed digital landscape. Over the next two days I received critiques of this work from René Ohlrau (@ArchaeOhlrau), an archaeology PhD candidate working on Trypillian “mega-sites” in Ukraine. As people who know me can attest, I have been craving a challenge to my work, not because I am spoiling for a fight, but rather because critiques from peers make me revisit my thinking to either change or defend it, making the work stronger (even if that means admitting that my initial thinking was either thin or outright wrong). Here then is a defense. I address the points raised in each of his four challenges in order, followed by a closing statement.

I.

“Why do you have to play the game for this? Why not ask the developer about their intentions to depict this area the way they did? There is no landscape development, because it’s artificial. Where is the archaeology in this?”

Response:

1. Why not ask the developer? I do plan on writing to Skyrim’s mapmaker, Noah Berry, to ask this very question, but it was important for me to do the work myself and to draw my own conclusions from my survey walks first so that my questions could have specificity and focus. I do not expect a reply, although I could be surprised. I have had some success with communicating with developers, namely when I studied The Talos Principle, and was able to get answers about Croteam’s use of Classical reception of both architecture and thought. I have written to No Man’s Sky’s developer, Hello Games, several times, but to no avail. To be clear, though, archaeology as well as art history and even literary criticism must get beyond the “ask the creator” stage. The work can incorporate their replies, but cannot rely on them completely. At times the creator can be an unreliable narrator, or have an unreliable memory, or can be short on time or frustrated and give an incomplete answer. Also, because the creator is so close to the work, other details about what was created can be overlooked or lost. The makers’ responses must be taken in the wider context of their work, but also with the elements surrounding and supporting their work. In a previous life I studied ancient Greek pottery, and I would have loved to ask Amasis why he painted eyes the way he did. I can’t, so I am stuck with his work and the context in which the work was created. That information provides me with a lot of data, incomplete without his biography, yet I can proceed with my archaeological and art historical investigation of his pots. With the archaeology of the recent past (or present or even future), asking questions of both creators and users is possible, but should be only part of the toolkit. Returning to the map of Skyrim, I was able to read an interview with Berry, and it bore out my observations, but I wanted to come to Skyrim’s landscape fresh, to see it for myself, and them compare what I learned to existing literature and words from the creator himself.
2. There is no landscape development, because it’s artificial. Where is the archaeology in this? I follow the definitions set by Edward Castronova that instead of “real” and “virtual” (or artificial), that we instead use “natural” and “synthetic.” In both the natural and synthetic worlds, landscapes are indeed developed, albeit on different timelines. As is always with the digital, creation is fast, certainly faster than geologic time. I would also argue that in-game landscapes do change. For designed worlds, these changes largely occur during the development cycle, but they also change (or have the potential to change) between versions. While this might not be immediately apparent in a game such as Skyrim, the landscape did change (and does change) in procedural games such as No Man’s Sky, sometimes by human agency, and other times by mechanical or code-based agency. As an archaeologist, I am interested in the oddness there is in static landscapes, those that do not undergo real-time formation processes. Just because we cannot observe a changing landscape, however, doesn’t mean that we cannot conduct landscape archaeology in synthetic worlds. Lastly, we must be careful in not confusing how landscapes work/behave in the natural and in the synthetic worlds. It is foolish to think that a designed landscape will behave as a natural one, but the same research questions can still apply to both kinds of worlds, and new questions can be developed about landscapes in synthetic spaces. Landscapes (either natural or synthetic) can be “built” or modified by people and other non-human agents. Understanding what is built and why and by/for whom as based on material evidence sounds like archaeology to me.

II.

“As a proof of concept, this is a nice idea. I still believe that an ‘Archaeology of virtual remains’ will directly go into the code to extract models like the pickaxe for example. As for the NMS survey ‘reverse engineering’, the algorithm might be the real deal?”

Response:

1. Archaeology of virtual remains: I chose to attempt to extract a 3D mesh from an artifact in Skyrim to see if it could be done, and if the procedure might map to other digital built environments. I play the game on the console, which does not allow me to find and use the “pickaxe” file as I could if I played on PC. I cannot access the code on my console, nor is Bethesda Game Studios likely to grant me access to their source code for anything. My goal was to place a virtual artifact on a turntable in a 3D digital environment to see if I could extrapolate data to produce a “real” 3D object via printing, turning an optical illusion of an imagined thing into something I can hold in my flesh-and-blood hand. I did this in preparation for doing something similar in a procedurally generated synthetic world where algorithms build new things all the time, that I could then extract via 3D printing.
2. Reverse engineering algorithms: I’m doing a number of archaeological things in No Man’s Sky, some of which intuits rules of that space through my embodiment, agency, interactions, material evidence, etc. Hello Games has mentioned their kind of uber-algorithm, which will never be shared with the public, and will not be shared with me even if I sign an NDA (should they ever reply to my emails). Sean Murray did share two video lectures about voxels and procedural world-building, which can inform how I look at the synthetic environment. But reducing everything to code neglects the archaeology of the procedurally created world itself, its randomness in creation caused by my presence in that space, not to mention glitches, which are themselves artifacts and contribute to the archaeological understanding of digital built environments. Speaking towards my original interest in archaeology, I could reduce a piece of pottery to the chemical composition of its clay and can then determine where that clay was sourced. I can understand typologies of the shapes of pots and derive general ideas about function and place of use. But again this is only part of the picture, which includes an artistic plan, a biography of the item, information on trade, possible burial customs, the identity of the maker and the users, a history of use, the context in which it was found, etc. I can learn much about a digital built environment by working with the material culture in it, most appropriately perhaps in something procedurally generated. But with whatever I find, it tells me a little about the environment itself, and about the maker, human or otherwise. My second PhD case study, which has passed its ethics review and has been approved by my committee, will find me conducting an archaeological investigation of the capital world of the Legacy Hub in No Man’s Sky, documenting human presence in an abandoned synthetic space. Here is a blending of the natural and synthetic, and I hope to see what questions I can both ask and answer based on what I find there. My third PhD case study will look at code itself, in this case the FORTRAN of Colossal Cave Adventure, to see what I can determine by treating code as a primary source, conducting a kind of code palaeography on something that is open source and authored by many hands over the past 30+ years.

III.

“It is about choosing the right method for the source. Right now you are like licking the stone to determine its chemical components. Restricting analysis to intentionally limited in-game mechanics will lead nowhere.”

Response:

1. Licking the Stone: A big part of the PhD thesis is determining the “right” methods for the source material. I suspect these will vary from environment to environment, and will vary even between different digital built environments. I anticipate that some of what I do will indeed confirm that the chosen method is “wrong”, but like any experimenter, failing saves others future futile effort. I expect to fail sometimes. Regarding “licking the stone,” that still gives me data about texture and taste and possible chemical composition. No observational effort is wasted, and there are many ways to observe and engage with phenomena, some haptic, some not.
2. In-game mechanics: I am not limiting myself to in-game mechanics of anything that I study. On occasion I will impose restrictions on myself so that I work within a set of defined parameters, which are often set as part of my research questions. Within any environment either natural or synthetic the archaeologist discovers rules (mechanics), some set by the maker, and others set by how matter behaves, and still others based on context and relationships between the observer and the observed. We can place ourselves within a game’s mechanics for some research, but can also observe things about the game from outside it. Both sets of observations contribute to understanding something with a digital presence. I can be in a landscape bodily to experience it sensually, but I can also observe the landscape from my desk through the filter of a map. Both observational methods give me data, data I could not get exclusively from fieldwalking or from mapreading.

IV.

“. . . especially for Skyrim appropriation by modders would capture social interaction with the media and not npc narratives placed in the villages.”

Response:

1. Modding. Examining mods and the modding community would add a lot of data to how humans interface with a non-human environment. For my case study, however, I am limited to my PS4, which to the best of my knowledge does not allow for mods or modding or for downloading and installing mods from places such as the Steam Workshop. Modders who adjust the game work with Skyrim just as formation processes change a natural landscape. People make a place their own. But the place must invite them in to make those changes.
2. NPC Narratives. In Skyrim the NPC narratives are canned and are often ridiculous especially when repeated as non-sequiturs. In a procedural game such as No Man’s Sky, however, speech and vocabulary vary. I suspect that future procedural games will feature algorithm-based languages with rules that create and evolve vocabulary, grammar, and syntax, languages that have never been seen/heard before, yet will be engaged with by many players. In Skyrim the NPC’s speech and narratives can provide information about a place, a kind of tavern archaeology within the synthetic world. It adds context to the material culture and to the landscape that supports it. While perhaps redundant in this context, it does train people to talk to the locals both for enlightened self-interest, and also to engage the population in a bit of public archaeology about their locality. When shared with the player community, the work involves them, too, and adds more context to the place being studied. Sometimes the value of doing something is not in the object being studied, but rather in the action of studying.

Conclusions

Digital games (but this can be expanded to all software) are perhaps the only things that can be classified as a kind of archaeological Trinity. Software programs are artifacts, sites, and landscapes all at the same time. I hold the artifact of a disk that contains a software executable and other files. I access the site of the software by running the executable. That program-site occupies a landscape of data along with the results of human and non-human interactions. It is both material culture and area of production for more material culture. I can change the program (or other human and non-human agents can). It evolves and experiences time. It is composed of arranged matter that behaves in different ways that can be observed either over the short- or long-term. It is archaeology, digital, something synthetic, and something natural.

After reading the work of Karen Barad, Donna Haraway, Bjørn Olsen, Manuel DeLanda, and Tim Ingold, I firmly believe that archaeology must take a post-human approach to understanding how people relate to things, how things relate to other things, and how matter relates to matter, both human and non-human. As humans, we tend to place ourselves at the center of the material universe, but we are no longer at the center, and perhaps we never were. That anthropocentrism prevents us from other ways of seeing, from getting to the heart of the matter.

—Andrew Reinhard, Archaeogaming