After observing an argument on Twitter about Anna Anthropy’s Dys4ia, I wanted to put some ideas down. Some game critics and theorists, most famously Raph Koster, find that Dys4ia does not offer what they expect from a game. Their response to this is to say that it, or parts of it, are not a game.

I like Anna Anthropy’s work, but I also try to be clear-eyed about the fact that a lot of Dys4ia could be built in PowerPoint and isn’t a game. That’s not a value judgement. My value judgement of the piece as a work of expressive art is pretty high.

Like Raph, I also really value and respect Dys4ia. I value it for the conversations it provokes, and for its landmark place in the landscape of trans-themed game development.


I personally don’t like it as a game experience. When I played Dys4ia many months ago, it did not give me what I wanted. I wanted to see a transgender story expressed in a form that I enjoy engaging with. I wanted familiar mechanics with an uncommon setting and story. Its hosting on Newgrounds, the number digit in the title, and the networks through which I became aware of the game all set me up with that expectation.

What I got was another form entirely. What I got was something that has very little in common with conventionally reproduced game mechanics and formats. For me, this was a disappointment. That was my basic, selfish, emotional reaction. I don’t consider that disappointment to even resemble a considered critical response. Dys4ia was not what I wanted it to be, but the job of game developers is not to feed me stuff I want all the time.

I learned later on that the game isn’t at its best when played by me. I heard that it is being used in educational contexts as a way of communicating about transitioning experientially. It is not necessarily for gamers. With an audience of non-gamers, those conventions and expectations are irrelevant and would be counterproductive. It is enormously positive that the player does not have to learn how to play like a gamer before engaging with the work. It is immediately and completely accessible. That is important for its educational value and it is a big part of the reason I value and respect the game so much.


Wherever we are on the spectrum, a big part of our negotiation with society about how it perceives our gender ends up being about boxes, boundaries and typologies. If your identity is non-conforming, then quite often it involves people suddenly and with startling frankness telling you where they believe you belong. That’s what it means to be a minority – people naturally slip into a power position that allows them to tell you what you are.

Dys4ia gets treated in the same way, with regard to its status as a game. Similarly, I think it’s possible that people only get to tell Anna Anthropy what her work is because they take on a powerful position against her. But in the same way that my gender identity is not about someone else’s perception of where I fit in, Dys4ia‘s status as a game or not is not about my experience playing it.

Tadhg Kelly shared in the twitter conversation his response to the charge that to judge something as not a game is to judge it negatively. In it, he suggests that instead, by clearly defining what a game is, developers might better understand what games can be. The implication is that it’s not derision that’s at work here: it’s science. On twitter, he said this is ‘not a patriarchal white man thing’


I don’t agree. I think that the exercise of defining something as ‘not a game’ comes from a position of power.

The problem with deciding from your own perspective that, universally, Dys4ia is not a game, is that it denies the validity of less vocal positions on that definition. I’m not even necessarily talking about any kind of oppressed identity, just any other cultural contexts that don’t turn up in the games industry discourse. Like games as TV show formats. Imagine taking similar pacing and mechanics to Dys4ia and using them on a game show. In fact, that’s pretty much what a lot of game shows are; a series of short interactive exercises. Compare Dys4ia to The Cube.

The game show example isn’t just frivolous. For the player who encounters Dys4ia as an educational tool, game shows could be a more relevant part of the semantic domain than RPGs and FPSes. For that intended user, it very likely is a game and fits all of their expectations of what a game should be.

By deciding that your frame of reference is the relevant one for defining Dys4ia, you’re taking an authoritarian position. This isn’t Tadhg Kelly’s intention; for him, it’s not about power but about a collaborative creative project in the games industry as a whole. But the activity of defining is part of what humanities people might call ‘a discursive power play that is often dominated by hegemonic positions.’

Words aren’t ever successfully defined by authoritarian voices. What actually happens on the ground is that people just use those words following how they are used by the culture around them. People use the words in connection with other words, and each connection contributes to the broader semantic field of those words.

You can’t change that reality by arbitrarily declaring a definition, but the power of your voice determines the influence of the connections you draw. Voices with less power will struggle to have the same impact on the semantic field, even if they are themselves the subjects of those words.


That’s not to say that there is no level on which you can come to an understanding of a word’s meaning without taking a position of power. Take as just one example a technique I learned at uni for trying to avoid letting my privilege dominate my analysis. Ethnologists have ways of analysing the use of words to understand their meaning in a foreign cultural context, such as domain analysis. Years ago, I made a comic application of domain analysis to Lady Gaga’s first album.

If I wanted to know what ‘game’ means, I would use a technique like domain analysis to understand its meaning in a target context. For example, if Lady Gaga’s first album was a text from the culture in question, this is what I would learn about ‘game’:

GAME, particularly a LOVE GAME, is something that both Lady Gaga and Cupid want to play. ‘You’ are ‘in the GAME.’ The GAME is an initial stage in the story of ‘us,’ along with a boy, a girl and a ‘huh.’

HUH is also something that you can put your hand on, and this co-occurs with smiling.

Which in no way resembles Raph Koster, Tadhg Kelly or Anna Anthropy’s understanding of the word. Nor does it resemble *my* understanding of the word, even though it’s a description that I wrote. That’s the important point – I don’t get to decide what a word means.

I wonder what we would learn about games if we studied them this way, but with less frivolous source material?

12 thoughts on “What is a game? It depends who’s playing

  1. I wish we had a different game to use as the poster child for this discussion. The fact that Dys4ia is itself about identity means that it adds to the political implication of every word and position.

    In the earlier article on narrative, I used Arkham Asylum as an example of the same craft element. And I think it’s important that those who object to the labelling realize that it is happening entirely within the context of a discussion on craft elements and techniques, and not even about the entirety of the game but of a PORTION of it. Within that context, the use of the term “game” DID arise from a domain analysis. A large part of the challenge is here is that the domain changed as the discussion evolved.

    Ultimately, though, for me the big takeaway here is that all sides should be willing to listen to each other respectfully, and engage in dialogue. That’s how boundaries are broken down, in the end.

    1. Thanks for your response. I hadn’t considered how rapid changes to the network of people and technologies might make semantic fields less stable than usual.

      I quite like the way certain games end up acting as prisms through which we can look at discourse and see how political it is. I strongly suspect that the act of defining always involves power in some way, and Dys4ia forms a useful prism for seeing that.

      1. Yes, I agree that defining always is an exercise of power. Sometimes, though, it’s done by craftspeople with the intent of gaining power over their medium, in order to increase *every craftperson’s* power.

        Now, intent doesn’t always matter, in terms of the political implications. But there are many domains where having precise definitions is of great assistance. Units of measure come to mind. Actually, all fields of engineering. And games unquestionably partake of engineering — differing games to differing degrees.

        I like the prism aspect too. But we choose to hold prisms to our eyes. Life would be greatly impoverished, I feel, if we had to stick to only one. :)

        1. I did almost include a bit about engineer-types vs. humanities types. This is an interesting discussion because it’s where engineers and humanities discourses end up rubbing up against each other in a very uncomfortable way.

          But my temptation to rely on seeing this as being about that dichotomy is an over-simplification. By advocating for an observation-based approach to semantics I’m taking the side of scientific enquiry over rhetorical discourse. And what you’re engaging in isn’t simply engineering, either.

          I think something is lost when we try to apply precision when the material resists it. Or at least, the friction that surrounds the attempt to define games is an inevitable result of the material’s slipperiness and its close connection to human subjects. To see how engineering-inspired approaches have negative effects when faced with human subjects, look at le Corbusier! You can design a freeway system with scientific principles, but you can’t engineer society quite as easily.

          1. Well, all my actual formal educational background is in humanities. Really, the closest analogy I can draw is that what I and other “game formalists” are after is identifying the equivalents of the iamb, the dolly shot, the golden section, and so on. In attempting to pin down why we say one thing is a sonnet and another free verse. It doesn’t mean that we then say “only sonnets are allowed!”

            Historically, I think that expansion of possibility has typically followed after formalism, and people say “this can’t be all!” and refuse to settle.

            So no, it’s isn’t really engineering. But prosody, harmonic analysis, and other such fields have similar precision goals, in terms of language.

          2. To be clear, I really, really love the work that is being done on smaller things that constitute games, and I would have been paralysed as a design historian without that work. I should be more careful when seeming to oppose formalism, because that’s not where I want to position myself in the discussion. But to answer your point more specifically, ‘what is a game’ is the equivalent of ‘what is a poem’, not of ‘what is a sonnet’. If someone was to write ‘this poem contains fragments of text that are not poetry’, that too would be heard by some as a negative value judgment. Moreover, analysis doesn’t require that the subject actually belongs to a particular form. ‘Let’s analyse this conference presentation as a game’ would be totally cool and I’m sure nobody would churlishly respond ‘you can’t apply that analytical framework, because it’s not a game.’

          3. I think we just hit the point where the terms start to matter for clarity of communication. I’ll use Lude as a replacement for Game for a moment.

            The first step of the formalism was actually examining the common elements within just the electronic game domain. From there, though, it broadened out to encompass everything that Wittgenstein recognized as having family resemblance. So in many ways, its first steps were inclusive: choosing to encompass sports, children’s play, boardgames, etc.

            Common elements were indeed identified, and from there the tag Lude was hung on that cluster. However, one of the most interesting qualities of this cluster was that it was recursive: a Lude was made of other Ludes. A given Lude could fundamentally best be understood as mathematical relationships modelling a dynamic system. However, a series of other characteristics were also present, not all of which were immediately evident: the role of an implicit opponent, for example; the fact that all Ludes are structurally sequential “turns” with time limit being a parameter that is often set extremely short (say, 1/60th of a second); and so on.

            However, gluing Ludes together commonly involved non-Ludes of varying degrees. These could be large, or small, in terms of their visibility to the audience; large, or small, in terms of their development effort; and large, or small, in terms of their contribution to the viability of the topmost Lude in the construct.

            There were plenty of cases where a Lude was NOT the topmost construct. The most visible example was MMOs, where the topmost construct we might call a World, and which then embeds Ludes. There were also numerous cases where verbs associated with Ludes (“play,” in particular) i a variety of languages suggested fresh avenues for exploring what a Lude might be. This was best articulated by Bernard Suits; that said, formalists went digging and found that invariably, ALL these activities that were described this way would have the qualities of a Lude imposed upon them in the process of “playing” them. In other words, audiences can spontaneously create a Lude through the dynamics of their interaction with a construct (mental, physical) that has a subset of the required elements.

            The fact that Ludes can be fruitfully extracted, diagrammed, modelled, and simulated completely outside of their context or indeed their presentation leads then to a wide array of possible tools. Among these is “whether the Lude is meeting its artistic goal.” Ludes are like machines, in some sense — they can be misdesigned, broke, produce incorrect outputs, or fail to plug into neighboring ludes in the appropriate way. So there is a craft element there where one can make value judgements about whether or not a Lude is doing what the craftsman intended.

            The important thing here, which I suspect is not clear to any but a small group of Lude grammarians, is that what we have of a Lude description is at this point

            – internally consistent (indeed, there are now tools to simulate them)
            – coherent
            – predictive
            – complete, in the sense that we have yet to find a Lude which does not fit
            – lexical

            Hence why the issue arises of whether “what is a Lude” is both “what is an iamb” and “what is a sonnet” and possibly even “what is a poem” or “what is poetry.” My own personal take is that the term fails to stretch to “what is poetry.” After all, for large chunks of my own career, I worked on Worlds, not Ludes. I would go further and add that there are other forms into which Ludes may be put (or echoes of Ludes suggested) which do not partake of the essential definitional qualities of a Lude. A random assortment thereof includes musical improvisation, participatory theater, some forms of storytelling play, a wide array of interactive artworks, and so on.

            In the spirit of erasing clear boundaries, many of these are termed “liminal.” Domain analysis on whether or not the word Lude applies in its common usage or as a layperson term tends to fail because the audience itself tends to be conflicted (cf Dear Esther, Wii Orchestra, Hollywood Medieval, etc).

            In common usage, any interactive construct that consisted primarily or Ludes was generally termed a Lude itself — not just in English either, but in a variety of other languages. The formalists chose to not fight that war, and simply went with common usage.

            That said, for lack of a better word for the thoroughly verified core qualities of Lude, the word chosen for within the problem domain was the word used most commonly in layperson communication, and which also accorded well with explorations in related fields such as mathematics. Within the problem domain, it effectively has a tighter definition, much like terms such as “mass” “rhyme” “harmony” and others do within their respective domains versus the general domain.

            I don’t know whether that serves as an answer to your post, but that’s where I am coming from. Once we remove “Lude” from it, it boils down to

            – a Game is made of Games
            – a Game may have plenty of non-Game in it
            – a non-Game may have Games in it
            – what we call “a game” in casual talk may or may not be a Game in the formal sense
            – what formalists call Game happens to line up pretty well with what mathematicians, sportspeople, sociologists, and philosophers have been calling a game
            – formalists are not interested in policing what laypeople call “a game” but are very interested in anything anyone says about what the boundaries of a Game are.

  2. Oh, a minor point: a big reason why Game ended up being the term rather than a new coinage is because there was an existing body of work that used that term linked with related terms such as Puzzle and Toy, which are central to understanding the entirety of the body of theory. So it isn’t even that the current generation of formalists made this choice… it’s a pre-existing one with multiple decades worth of usage.

  3. Wow, having an opinion about her work is not “[taking] on a powerful position against her”. If anything, Anthropy’s angry refusal to allow others to interpret her work in their own frames of reference is the problem.

    1. That slightly misrepresents what I wrote: I very clearly said that I think that, specifically, the exercise of attempting to set universal definitions can only come from a position of power. I have lots of opinions about Anna Anthropy’s work, I don’t think there’s anything wrong with that, but opinions are complex things and sometimes our opinions have as much to do with our own assumptions about our position in the world as they do about the work itself. Being a person with opinions is complicated if you are aware of social context and power dynamics.

Leave a Reply

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