Digital resources in the Social Sciences and Humanities OpenEdition Our platforms OpenEdition Books OpenEdition Journals Hypotheses Calenda Libraries OpenEdition Freemium Follow us
Photo of Mercer Island Arcade 1982, © Natalie Fobes/Seattle Times

What does it mean for a game to be a source?

by Tobias Winnerling

In June 2025, the CH Ludens team had the pleasure to welcome Tobias Winnerling (Heinrich Heine Universität Düsseldorf) for a guest lecture at the University of Bern as part of our yearly workshop. This year’s workshop, organized by the Bern team of CH Ludens, focused on historical source criticism and the problems and possibilities of historical sources such as games, gaming magazines, and oral history interviews, and we started with said guest lecture, titled “Digital Games, Digital Born Sources, Discourses and Dispositifs”, and with this blogpost, we are publishing the written version of the first part of this talk.

Historians like to point out that their writing is based on verifiable empirical facts. They make extensive use of ostensive references, to speak with Jean-François Lyotard: Pointing to specific, distinctly identifiable material entities from the past outside of the historian’s texts which are called upon to authenticate the statements in the text. Within an historical argument, these material entities are called upon to supply empirical evidence corroborating the interpretations. And, inversely, they act as delimiters for the scope of historiographical statements: You cannot put forward any historiographical claim for which there are no historical materials to support it. If there is no evidence from the past in question, we cannot reasonably say something about this past.

Traditionally, these specific, distinctly identifiable material entities from the past are called “sources” by historians, and although this term itself is not unproblematic, I am going to use it for all kinds of historical material in certain relations in the following. Over the last three centuries, historians of all kinds have developed quite sophisticated procedures for identifying, collecting, preserving, filing, and processing sources.

But first things first: What does such an entity from the past need to become a source? Well, it must be useful in contributing to answer a specific question about the past – and thus to help to gain what we like to think of as knowledge about the past. From this simple observation two important things follow: First, that nothing is a source unless being used as such; there is no such thing as a historical source in itself. Something is a source if and only if it shall contribute to answering a  historiographical question. Sources are always “sources for X”, X being something you want to know about the past. And second, it follows that any remnant of the past can be turned into a source, only depending on the respective historiographical question. Anything left from the past can answer certain questions about the past, even if it has been disposed of as waste centuries ago.

What’s in a game

That in turn means that also every digital game from the past may be turned into a source about the past: it all comes down to the specific historiographical question. Well, at least theoretically, as there unfortunately are some problems to be faced in the process. The first is to determine what actually constitutes “a digital game” as a distinctly identifiable entity.

That’s already a tricky one. As the common consensus among researchers in the field has it at the moment, a digital game is a complex yet highly ephemeral phenomenon emerging from the interplay of three interrelated factors or components: A player component (usually human, but may also be an artificial agent such as an AI), a program component (containing rules and parameters, in a digital context the respective software), and a technical component (the hardware necessary to process both the program component and the player’s inputs).

Graph from Tobias Winnerling's presentation: Triangle of emergence: Digital Games as phenomena
Ill. 1: Triangle of emergence: Digital Games as phenomena

The game in its phenomenal entirety as – all too briefly – sketched out above could be regarded as something that may be distilled into a single distinctly identifiable material entity only with great difficulty. But only then it can be safely stored in an archive, filed, described, and from then on called upon as historical evidence if need be. And this is how sources are handled. Regarding the player component this difficulty becomes obvious, as neither the players nor their actual, situatively emerging play can be collected as material entities of evidence to be filed away into an archival box.

This is a familiar problem with textual and non-textual historical sources alike: all of these have been made by men, most of the time for being used by men, but usually neither the makers nor the users of the past are available for us as historical evidence. That in turn means – where there are no sources there are no reasonable historiographical statements – that some aspects of the past, especially concerning everyday practices, can only be made the subject of fruitful historical research with great difficulty if at all (and most of the time only indirectly by way of other sources or material traces in the objects). Standard historical procedure most of the time would be to just ignore the users in terms of material source objects.

Following this line of thought, one may come to the conclusion that for the same reason the player component needs to be left out of discussing the digital games as sources. We would be left with the program and technical components as the two remaining parts of the triangle (Ill. 1), as those may be collected as physical objects or on physical objects as storage media. This would leave us with a potentially quite complicated array of physical electronic devices and – at least in theory – virtual software, all of which would need to be filed away very carefully in order to maintain its functionality. This solution would involve quite considerable and rather strenuous efforts. Isn’t there a less troublesome way? Maybe yes. If the player component may be disregarded as an essential part of the distinctly identifiable individual entity “digital game” to be designated as a historical source, perhaps the technical component might be disregarded, too. That would save a lot of work and worry, and there actually is an argument to support it. This argument runs as follows: There is no digital game without software; software has no necessary connections to physical hardware, as it may be processed via emulation and thus be run on a multitude of different devices; therefore, to identify, collect, preserve, file, and process “the digital game” as a source, it is sufficient to rely on the program component alone. In such a view, the game would effectively be reduced to pure code, and everything else relegated to the status of merely accidental peripheral materials.

Datawise, this is actually the easiest way to manage digital games as sources: They may be saved and archived quite easily like that. The problems arising here are of another nature, having to do with the fluidity of the program, which may be steadily patched, updated, and added onto. The issue has come to prominence by the example of MMORPGS, where the discussion started at least with the first „vanilla“ World of Warcraft servers being set up to preserve a certain state of game world and game in the early 2010s.

Yet, the fact that it would be theoretically possible to reduce games to their program component to define them as historical sources does not say anything about whether that would be a good idea at all. There are specific problems connected with this approach apart from sorting out versions, patches, bugfixes, and mods as just detailed. These problems touch a more structural level of research issues. As the reduction of digital games to their program component is the narrowest definition of games as historical sources possible, it is subject to both the structural advantages and disadvantages of narrow definitions.

The advantage of a narrow definition is of course that it is precise, clear-cut and easy to apply. The corresponding disadvantage is that narrow definitions exclude a large part of the historical material at hand, as it will not fit into their precise and clear-cut scheme. Thus, they foreclose possible research directions and questions if these cannot be answered by relying on the constricted set of sources made available through the definition. So, the gains in definitional precision are to be balanced against losses in possible perspectives, and it may easily be the case that the latter outweigh the former.

How to complicate things

Photo of Mercer Island Arcade 1982, © Natalie Fobes/Seattle Times
Ill. 2: Mercer Island Arcade 1982, © Natalie Fobes/Seattle Times

Let me illustrate this by a short example: The picture above (Ill. 2) was taken on February, 10, 1982, at Mercer Island Arcade in the Greater Seattle Area, USA. It shows the typical video game arcade setting of the early 1980’s: A lot of arcade cabinets lined up against a wall with a crowd of mostly young people in front of them, some playing and some watching. Most of the people depicted seem to be engaged in some kind of social interaction. For the following I want to focus on the two (presumably) boys in the centre of the image who are playing what is easily recognizable as a Defender cabinet. Defender only being released in 1981, this was brand-new entertainment at the time.

It’s interesting to see them both at that particular cabinet as Defender, although acclaimed as a digital game classic nowadays, did only slowly catch on at the beginning because of its complicated controls. Each player had to operate one joystick and five buttons, and, with the limited space available both at the cabinet and in the arcade setting, in a two-player game this became quite challenging because there physically was the need to shuffle together somehow and still be able to manoeuvre.

This offers the opportunity to ask one seemingly simple question: What exactly is the digital game that could be termed a historical source in this configuration? Judging from what has just been discussed, there are three options. The first would be to restrict the game to the program code that was necessary for the arcade cabinet visible in the picture to process it and to let a gameplaying session of Defender emerge. That this would indeed be conceivable is supported by the many Defender emulations available which since 1981 have ported the game to quite a lot of different platforms so that the technical component may be seen as accidental here.

But on the other hand, all of these different versions really are different versions in the sense that they do not reproduce the controls of the 1981 arcade cabinet. You do not have to navigate the complicated joystick-and-five-buttons configuration which earned the game its early reputation for presenting quite a challenge to its players. And, as a perhaps literally more substantial argument, it was precisely this configuration of controls which was purposely built into the cabinet in co-evolving accordance with the program. So, the controls were not only hard-wired into the cabinet, but it also still is up for discussion which of the two was the dominant factor in determining the ultimate outcome that resulted in a Defender cabinet to be lined up with others in a large arcade hall. Most probably we will never know whether the physical controls or the virtual program code were decisive here. That would be a good argument to choose the second option and to say that it is precisely the combination of program component and technical component that constitutes the source here.

But as the actual situation visible in the picture is not only determined by the features of program and cabinet but also by the fact that the two children at play here shuffle for space in front of the cabinet, with what most probably are their school bags still on their backs, further restricting movement and demanding more space – which would have presented a more pressing problem would not luck have had it that both cabinets to the left and right of them had not found any players in that particular moment – there also might be a valid argument to say that a discussion of the digital-game-as-source here should comprise the social component also.

On top of everything any sharp observer, let alone one with a bit of historical expertise, will  have figured out at this point that what I am talking about is a quite different source as well. The photograph is the only document surviving from this gameplay situation. Moreover it may have been staged altogether as it was part of a report about Mercer Island Arcade gaining scrutiny from concerned local parents that it would be the kind of place they did not want their kids to meet at. It definitely is not part of the digital game Defender which has been discussed up until now, but it is part of a media ensemble which revolves around this game, or, more precisely, these games as there are seven cabinets visible in the picture. Thus, this related media environment needs to be factored into the discussion in some way or other as well.

Multi-Layered Environments

Graph from Tobias Winnerling's presentation: Layer model of Digital Games
Ill. 3: Layer model of Digital Games

Therefore, I would like to propose a typology and terminology that allows to precisely indicate all these definitional levels of description or analysis when talking about digital games as sources (and, hopefully, it may be useful even in the wider context of scientifically handling digital games). This typology starts out from the narrowest possible definition, that is, the definition solely via the program component of the triangle of emergence. I take this as the basic layer and first level for subsequent definition efforts. In the typology, to avoid confusion, I have labelled this level as the Program Arrangement, consisting of the program, rules and other necessary non-material factors determining the matrix of the game in question. In the example this would correspond to the Defender software.

For a wider discussion, the next step would be to change to the second layer of possible definitions, consisting of the technical setup the Program Arrangement is embedded in, which would apply to the arcade cabinet of the example. This layer of the material technical level I therefore propose to label as the Technical Ecosystem of a given game.

Moving wider into the open, the tertiary layer then consists of the Social Ecosystem the Technical Ecosystem and Program Arrangement of the game are both embedded in. Relating to the example once again this would be the level of discussion that includes the two children playing at the same arcade cabinet and shuffling for position at the controls, and maybe also the social setting of the arcade as a specific (and special) location.

And to also incorporate the photograph and other related media, these would form a tentative quaternary layer which I propose to label the Paraludic Ecosystem of a given game consisting of all kinds of game-focused media, technologies, and practices which are not part of the game proper but which form another, even wider ecosystem of the game into which the other three are embedded.

From layer to layer, the scope of the phenomena encompassed under the focus of the digital game widens, which multiplies the potential questions to be asked. At the same time, the delineations of the respective ecosystems become ever more fuzzy and blurred from layer to layer, which decreases the exactitude of the definition of what is to be taken as a source (or a corpus of sources) in this context. Specifying on which level of definition the respective analysis operates at a given point in time that we are studying or in a specific part of the text we are writing  thus serves to point out and clarify the extension of the terms we use. The possible scope and degree of exactitude that can be expected of the conclusions would then be clear since both depend directly on the level on which we are. Hopefully this would also raise the degree of precision and unambiguity of subsequent discussions.

Gaming the Definition

Unfortunately, all of this offers in no way a definitive answer to the question posed above: What does it now mean for a digital game to be considered as a source?In fact, it is by no means obvious which parts of a given game should be considered to constitute the game as a source. We will always need to construct these source entities by definition, and most probably will not find a single definition matching all cases. So why not  approach the question from another perspective? If we see the different components as layers of possible extensions of any conceivable game, we may flexibly decide how many such layers we want to incorporate into our notion of „the game“ as a source, and yet be able to record this precisely. Instead of asking „what kind of source is a digital game?“ we could ask „What kind of source must a digital game be for the purposes of my research?“

This is hardly original: In dealing with manuscripts and early print books as sources, historians of all kinds have used similar approaches for centuries already by distinguishing an ideal text as content from the individual physical copy and its individual material text.

Comparison: Early printed books as sources
Ill. 4: Comparison: Early printed books as sources

This results in the same and still unsolved problems of differentiation between both variations of the text and their respective interrelation in constituting the source of a particular text or book, but it has not hindered historians to base their arguments on either or both for some centuries now. By way of comparison, it serves again to draw attention to the fact that most of the problems research on digital games is facing are in no way new, nor are the solutions devised to cope with them. When encountered in the context of more traditional media we only tend to overlook them as they have been silently neutralized in the everyday practices of research. This has most often happened by not asking questions about them anymore, as these questions have been determined too difficult or too unrewarding to resolve (or both).

Yet, there always are disciplines taking care of such matters. Anything I said about the secondary layer, the level of the Technical Ecosystem, would sound utterly familiar to a researcher in Book Sciences or Codicology. In the same way, what I mentioned about the tertiary level of the social ecosystem would in no way surprise any cultural historians or philologists concerned with what they call “writing scenes” or “reading scenes”. My remarks about what I termed the Paraludic Ecosystem as the quaternary level would sound perfectly self-evident to any researcher in the media or communication sciences, to pick only a few examples for each layer and level.

So, when it comes to viewing digital games as sources, the most basic task research has to face is that they really are not that special on the one hand – and that they are special enough to merit a reasonably precise self-reflection of the level and scope of the respective analysis and its terms by their researchers on the other hand. That should be the rule of the game.


Header Image: Mercer Island Arcade 1982, © Natalie Fobes/Seattle Times.


OpenEdition suggests that you cite this post as follows:
CH Ludens (September 22, 2025). What does it mean for a game to be a source? Confoederatio Ludens. Retrieved November 15, 2025 from https://doi.org/10.58079/14q24


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.