This is the fifth article in the series The Academics Are Coming.
Dr. Clara Fernández-Vara is a researcher for the Singapore-MIT GAMBIT Game Lab and much of her research is focused on the integration of story and gameplay. She has been involved in the development of several games through her studies.
She’s written an interactive fiction, Ariel, based on Shakespeare’s The Tempest, but the bulk of her work concerns point-and-click adventuring. Rosemary (2009) adopted the “remembered past” as a mechanic (previously featured on Rock Paper Shotgun). Symon (2010) explored procedural generation in the adventure space using dream logic to sweep any narrative problems under the carpet (flagged on RPS just yesterday). The follow-up to Symon was Stranded in Singapore (2011) which aimed to standardize the procedurally-generated approach.
Fernández-Vara talked to Electron Dance about her work, what she loves about developing academic games and how research should not be constrained by the concerns of industry.
HM: Clara, there is still much scepticism regarding the value of academic research so… why become a game studies academic?
I was already on the academic track; I started my PhD in English and American literature in Spain to study Shakespeare on film. At first it was a secondary interest, tracing the similarities between traditional fairytales and videogames, particularly arcade games. Literature gave me the excuse to rediscover games, because literary theory helped me understand them in a new way.
Then I came to the US to study media studies at MIT, where I was surrounded by people studying comics, Bollywood movies, radio production; games were one more topic in the mix. I also started to hang out with people like Jesper Juul and my colleagues at the Education Arcade, who were games researchers and developers. Talking to them, I realized that there was so much to learn, that game studies was a vast field, mainly unexplored. My background in literature, and mostly on studying texts in performance, had provided me with good tools to gain important insights. I felt like I could make a contribution, which is essential to be a scholar. Making games was also very attractive–if I was part of theatre productions as a theatre scholar, it seemed natural to participate in production to understand my field of study better.
I’ve realized I just changed the subject of study, from theatre and film to games, and brought all my methods along. I focus on story-driven games because of my literary background, and work on adventure games because it is one of my favourite genres, and I know how to design and program them best. So basically I wanted to stay in academia, having a subject of study that was challenging and engaging. Although I have less resources than I’d have working in the industry, staying in academia gives me the liberty to make games about themes and using mechanics that commercial games would usually not tackle.
As for some people’s contempt towards games scholarship, it’s not something I’m too worried about. One of my teachers in Madrid frowned on me when he heard I wanted to study games, he said it was “a waste of time” and that I should stick with Shakespeare. I think he was one of the reasons why I study games–I wanted to show him and others that games are not only a very valid field of study, but that by understanding them better we can make the expressive possibilities of games thrive. Plus I’m a bit of a rebel–whenever a teacher told me I should stay on the beaten track to succeed, I promptly ran the other way. So far, it’s worked pretty well for me.
HM: How would you define your role on Rosemary, Symon and Stranded in Singapore?
As a researcher working with student teams, my role is first explaining what my research is and what I’m trying to learn from making the game. I guess that in industry terms I’m closer to a creative director. I start by explaining my design philosophy to the team, and my focus on exploration and world building, on letting players discover the world and make sense of it. There are certain design choices that I usually make beforehand, such as what type of adventure game we are making, or the topic of the game, e.g. the core mechanic is remembering, or the game takes place in Singapore. The students have ownership of the game too–it’s not made on spec, but rather is the result of the collaboration of the whole team. I’m leading the orchestra, but I cannot play alone.
Rosemary was the game where I had the most hats–I was the producer, and also contributed to the design and writing of the game. In Symon and Stranded in Singapore I was a bit more of a client and had a game director (Jason Beene for Symon, Rik Eberhardt for Stranded in Singapore) who helped me interface with the team. It makes my life a bit easier, because they dealt with things like scheduling and mentoring the students on game development practices, and allows me to focus on the design of the game and how it addresses what I want to achieve. But I love working with students, so in both cases I still worked closely with the designer and the rest of the team.
It is always tricky working with teams of students, because on the one hand the game is supposed to help me with my research, but on the other, they are learning. They are going to make mistakes–and they should be allowed to make mistakes every now and then. I have not completely figured out the process, but I have to say I love the energy of working with undergrads. I need to get my hands dirty along with them, because I also learn a lot from the team.
It’s also a great way of putting my theories to the test, since I have to explain my philosophy and research to them and make it clear.
HM: What was the motivation behind these projects? Were they experiments or demonstrations of theory?
They are definitely experiments because the goal is exploring the process. Even when the game or prototype does not quite come together, I still learn from making it. My theories are frameworks to understand games and to formulate specific problems; at times the goal of the game is to subvert them and put them to the test. The validation comes from releasing the games to the world and getting people outside the academic realm to play them. My goal is making games that are different, novel, and engaging; having people play them without realizing that they are research games shows that what I do is relevant and useful.
My general goal is to experiment within the adventure game genre, questioning preconceptions and looking for ways to innovate. I focus on them more as games, and how to bring together gameplay and story. Rosemary was a general experiment–at that point, I wasn’t sure we could make a short adventure game with the resources we had at hand. The experiment was bringing new mechanics to adventure games–could we make a game where the core mechanic was remembering? What was the process?
Symon is an example of a design question that can take you to findings in a variety of fields. The premise at first was questioning my own theories–I argue that narrative puzzles have to be set up in a world that is consistent, that creates expectations, so that when the player encounters the puzzle, she can solve it according to the logic and affordances of the world. For example, in the game Loom the player interacts with the world through musical spells, which are learned through exploring the world. So I poked at my own theories: can we make a game based on a dream world, an unstable place where the logic doesn’t quite make sense. We all know the feeling–what we do while dreaming makes perfect sense to us, but sounds crazy when we think about it what we’re awake (if we remember what happened).
After some experimentation and prototyping, one of the approaches that seemed the most promising was creating procedurally generated puzzles, first because it would provide that sense of instability. It was also a promising design solution: the puzzle generation system would have to model the mind of whoever’s dream the player is in. It also opened up a really exciting avenue for innovation in adventure games, since procedural generation would make the game replayable. There were several prototypes on the topic as preparation and Symon was the demonstration of how it would work.
From Symon, we learned that procedural generation created a new type of adventure game: shorter but replayable, where the player figures out the world through several playthroughs. The problem was that the designer did not have good tools with which to create a procedurally generated adventure game and it took very long. So the next step was developing those tools and putting them to the test making another game. For that game, I posed an extra design problem–the randomness of the procedural generation works well in a dream setting, but could we use it for games based on the real world? Otherwise, our tools may only be good to create dream-like games.
The process of development of Stranded in Singapore addressed that question: the tools worked quite well but the design, however, was more difficult because we were simulating aspects of a real place (Singapore) and not everything fits in a procedurally generated design. It was lucky that food is an important part of Singaporean culture, and that fit well, but it was a hard constraint, and not all of the puzzles work well.
In spite of this, my focus is not on procedural generation, but rather in the conceptual design. One of my students is extending the development tools for his Master’s thesis in Computer Science. My research focuses on finding the problems and creating solutions, and reflecting on the process–if in the process there are other pieces of research that can advance other fields, like computer science, that’s great, it’s the beauty of working in an interdisciplinary field like games.
What I’ve learned too is that as a researcher it is easy to constrain yourself too much, and that what may be good research may not result in the best game and vice versa. On top of that, teaching is also part of the process. Right now, I’m rethinking my methods, trying to figure out what fits my background best, what is the most satisfying process professionally, and how can I learn the most. I may have been trying to be too scientific about it and I’m mainly a humanist. So I’m looking at models of theorist practitioners in other media to learn more about their process, such as Eisenstein or Truffaut in film, or Anthony Burgess in literature, to look at how they combined their theories and criticism with the works they produced.
HM: In your view, how much impact does games studies have on the world of game development?
Not all of game studies work is going to be relevant to industry and it does not have to. As academics we need the liberty to explore our field. Research does not always have to be at the service of industry. From the humanities standpoint, game studies helps us understand games better, and some of those insights may be superfluous to developers. Same with technological research (a novel technology may not be feasible to commercialize) or the social sciences (studying defunct online communities) where the research is excellent and advances knowledge, but may not have an immediate impact outside of academia. I have chosen to make my work relevant to practice but not everyone has to, and I’m well aware that only some aspects of my work can be made useful to developers.
My friends in the industry very rarely read papers from academic conferences or journals, mostly because they do not have time. Papers are not the most efficient format to influence the industry, so we have to find other ways to communicate. One way would be having some more academics presenting at industry conferences. Until a couple of years ago, there was a GDC session called Game Studies Download, which provided a selection of papers that were relevant to the industry; we need something like that back, for example.
As academics, we need to realize that if we want the industry to hear us, we have to make our work accessible to them. We can write a paper for Game Studies, but we can also have a more accessible version in our blogs; not dumbing down our work but rather being careful with how we use jargon, for example, or thinking about how our research helps understanding games better. We cannot go into a room of developers and read a philosophical debate on the misunderstandings about Huizinga’s and Caillois’ work, which makes for a great session in a research conference like DiGRA, but would reaffirm the general scepticism that the industry already has towards academics. Making games is another way to gain some respect from industry–if you can link your theory with your game, it’s more likely that they’ll pay attention. Developers may not read papers, but they do play games.
I believe that the influence of game studies in the industry is barely visible now, but it’ll be more prominent as the field grows and more game studies academics teach people who later go into game development. It’ll also be more patent in smaller companies than in AAA companies where the contents may still be more controlled by publishers and big money. So I believe we’re getting there.
HM: Could you be tempted into larger projects with more commercial aspirations?
I keep saying that if I can deliver a game in eight weeks and a team of students who are new to development, then I’d be on fire if I could get to work for four months with a team of professionals. My goal is to be relevant, so applying what I know to larger commercial products would be a fascinating challenge and I would learn a lot from it. I’d love to have access to more resources, and face the challenges of bigger games.
One of the first things that I have to think of when making a game is my target audience, because they are the ones who will be making sense of the game. So far my goal has been to appeal to relatively niche audiences; reaching out to more people would be an interesting problem to tackle. I’d love to make a game that reaches larger audiences, and see what people do with them–that’s the biggest kick I get from making games as part of my work.
On the other hand, I’d be more worried about working on a commercial product where my creative input is exclusively at the service of what some marketing department or team of executives thinks would sell, making clones of other successful games. So making larger experimental games for the market may be a fun challenge; a humongous game cash-in is not worth my time.
HM: Thanks for talking to Electron Dance, Clara.
Further Reading
If you’re interested in browsing through Fernández-Vara’s academic output, you can find it available online. Her most recent paper was Games Spaces Speak Volumes: Indexical Storytelling on refining environmental storytelling as an integral part of gameplay as opposed to just back story.
Download my FREE eBook on the collapse of indie game prices an accessible and comprehensive explanation of what has happened to the market.
Sign up for the monthly Electron Dance Newsletter and follow on Twitter!
Interesting – I read about Symon on RPS and decided it didn’t appeal to me, but I might have to take another look. An adventure game sacrificing strong authorship for the sake of replayability seemed rather to be missing the point (similarly I wouldn’t be enthused by a mystery novel where the culprit was randomly determined with each readthrough). Randomly generated elements worked well for Blade Runner because they created a genuine sense of uncertainty and the sense that it was fairly arbitrary which of the characters were replicants actually matched the theme of the game – that doesn’t work for the general case. But if there’s a point to the replayability itself – piecing together a bigger picture when you only see part of it each time – that sounds more interesting.
I haven’t had a chance to read the Indexical Storytelling paper yet but I’m looking forward to it – would I need an account to download the pdf version?
Thanks HM and Clara!
Hi phlebas. I’m not going to say “everybody must play Symon” but I found it interesting enough to play several times to see how it worked. I’m not convinced myself of the procedural generation in the adventure space – Symon and Stranded seem to reduce the point-and-click largely to a consistent randomisation of fetch quests with a thread of narrative relevancy associated with these quests. I’m not sure that’s where we want to end up.
On the other hand, we all have to start somewhere. Baby steps are necessary before you can start tackling more serious issues. I’m pretty sure the procedural generation ground covered here was no cakewalk and full of little pitfalls that weren’t expected. I guess ideally we’re looking for something as complex as Façade.
The PDF is free to download no account necessary. Clara references Demon’s Souls as an example of indexical storytelling – remnants of other players’ activity are environmental narrative that are actually instructive to your gameplay.
If anyone isn’t clear on the term “environmental narrative”, here’s a comprehensive piece in the Electron Dance archives.
Hmm. I can read the paper onscreen from that link, but the download button tells me I need to be logged in. I wonder what I’m doing wrong.
Some of those other papers look promising too, especially on puzzle design and innovation.
My bad! Looks like it expects either a local account or Facebook. I was logged into Facebook at the time of my download and didn’t get asked. Icon suggests Twitter login will also work.
Ya know, I’ve been thinking about procedural generation for my next post. I’m not convinced it’s all that useful in some contexts. Everything doesn’t need to be Dwarf Fortress.
Anyway, this is what I’ve been looking at. Have you played it? Symon looks like it might be a good point of reference because I think the interactive fiction and adventure game spaces share some overlap.
Outstanding interview. I’m really enjoying this series. Haven’t checked out any of Clara’s work yet, but I’ll do so today. Keep them coming!
@Alex: I’m extremely happy with the concept of an author telling a story in games. There’s a school of thoughtwhales that believes this is old media, that games need to move away from that into something more freeform. We need to stop writing linear stories and wrapping game around them. Personally, I have no problem with the throwaway game – play once vs play forever. (Funny how we head back to our discussions during A Theoretical War: the idea that games should be infinitely replayable, one of the key arguments employed against story-based games.)
I’m a little torn about these text-based choose-your-own-adventure-book experiences. I can see what’s going on here; Rob Gilbert’s point-and-click innovation removed the problem of parser guessing (see David Fox’s interview) but also threw away the idea of text. So what about the hybrid? Keep it text-based but remove the parser problem by reducing it down to branching narrative, which is what most IF boils down to, anyway. Yet: I am an old-skool romantic. I think the action of entering commands that you thought of without being shown a list of possible actions is an important part of the experience. Being handed a “road map” of the narrative just doesn’t give me the same buzz, decaffinated text adventures. So I’m not a great fan of Twine games either.
Saying that I didn’t play what you linked to, yet =) MY ENTIRE COMMENT MAY BE NONSENSE AND THE INTERNET WILL CROWDSOURCE THE HORRIBLE TRUTH
@Steerpike: Cheers Steerpike! Clara’s interview is a perfect example why this needed to be a series; I only asked a handful of questions and her answers were so detailed that I didn’t really need to ask much else. The same thing happened with most of the interviews… I just couldn’t bring myself to squish all these thoughtful responses into a single article.
Excellent read, I can see why you wouldn’t want to squish this down, there are far too many interesting things being said. Thanks for this HM and Clara. Now if you’ll excuse me I’ve got some of those links to check out…
HM: If you want text-based point-and-click that isn’t a CYOA, you can try A/The Colder Light. It takes away the parser guessing by having a bunch of verb and noun buttons, but there are enough combinations that it’s definitely not just a road map of the narrative; it’s a text adventure like you know, but with a different interface for entering commands. Which means that the action of entering commands that you thought of is still gone, but it’s still worth checking out.
Matt – I should’ve known someone would have done this already! Looks like it’s been built atop the Inform system.
Great interview, HM and Clara both.
The bit about procedural generation in adventure games got me wondering; we know the narrative thread is wrapped around what items the game decides you need this time, for replayability’s sake. But perhaps the interplay of the items and story would seem more worthy of analysis if we didn’t know it was somewhat random. Even if Minecraft generated places that fit the environmental storytelling bill, we wouldn’t ascribe the same value to it since we’d know they weren’t done so on purpose.
Is procedural generation something best kept secret? Is the next step coming up with ways to hide its use (not just in appearing handmade, but in the fact that it’s done on the fly)?
Another question, more specifically for Clara (unless someone else knows). I noticed that all of the games shown have lovely art. Was the style decided in concert with the concept to evoke it more effectively? I know when it comes down to pulling a game from the sea of indie platformers, I give preference to ones with art I like.
P.S. Clara’s picture matches the post’s title perfectly.
Funny you should say this Sid about “knowing about the randomness” because I make a spookily similar point in tomorrow’s post about another recent title. You’ll know it when you see it.
In contrast, that’s why Angelina has attracted interest, because it is trying to make connections with its procedurally-generated content (although I’m going to hold off giving it a thumbs up, because I’m not sure how innovative it can be from a ludic standpoint).
Hiding the procedurally generated nature of a game is going to be impossible. If we attempt that, some are going to be gutted when they find out they’ve been “fooled” and that there was no point in what they perceived as clever writing. We already have this problem in authored games where scripting is meant to make us feel danger when, actually, we aren’t. We don’t like it when the game has lied to us.
I suppose that’s true. Maybe designers could just procedurally generate a world, write around it, and ship that seed as “the one”.
Obviously if they still included the generation method, the writing would fall apart, but I’m awfully tired of games trying to tie full-quality presentation to obviously ancillary features anyways. That just results in less of them.
Ahh, now that’s interesting. That statement sort of comes full circle to my view of artists throwing random or hand-wavingly vague things into a pot and leaving the audience to make what they will out of it all, though at least with procedural generation the soulless computer or code can always be held accountable. Unless an artist admits ‘Hey, I’ve no idea why I did that and what those things really mean or how they fit together’ (which is highly unlikely), then they can get away with their mode of generation scot-free while everyone keeps interpreting away as if there was intentional meaning there all along. An artist being simultaneously gnomish and keeping their cards close to them and revealing nothing, and a computer ‘secretly’ (for as long as it remains a secret) procedurally generating things would surely yield the same results (whether you’re okay with being ‘fooled’ or not)? The difference being you’d never really know whether the artist ever meant anything in the first place while the poor computer is going to get rumbled sooner or later.
It would be interesting to put three of Bogost’s best machined haiku’s beside three ‘real’ haikus then letting a bunch of people discuss them. Later on you could reveal the truth about the machined haikus and see how that then affects their views.
I’m surprised there’s very little on the web written about Muriel Gray’s Art Is Dead – Long Live TV which was aired on Channel 4 in the early 90s. It caused a big controversy at the time. Gray interviewed four rising artists who created the type of modern art that was coming under fire as “random shit dressed up as art” by the general public and defended by critics. We were in the wake of Tracey Emin and Damien Hirst. One of the artists made sculptures made of vomit and rotting flesh.
Gray, if I recall, was fairly disparaging about the artists and some critics leapt to their defense. In the final episode of the series, she revealed it was a big hoax. None of these artists existed. And the critics defended them.
And unexpectedly we crash straight into the modern art debate. Thanks, Gregg.
David Lynch. Over the years, I’ve got the impression that he doesn’t knows why he is assembling various surreal scenes but has a gut feel that This Is Right. That works on films but can be more tricky for a long-running serial like, er, Twin Peaks. There’s a narrative intent but probably less strong than something like Forrest Gump. It’s almost procedurally-generated, a sense that all this works on some level but it’s the audience’s job to work that out.
So the problem I think we’re hitting here, which I guess comes also comes out in today’s Dear Korsakovia piece, is that random shit dressed up as art is sometimes indistinguishable from artistic intent.
I guess, Gregg, if the computer comes up with something that makes thematic sense – time and time again – we’ve basically got ourselves a robotic author. But I am troubled by the idea that the player can be expected not just to fill in the blanks but, actually, make up the whole story; it seems regressive somehow, a way to pretend that machines can’t make stories. This isn’t a criticism of Symon, Stranded in Singapore nor Dear Esther: I’m just thinking out loud about where some of these developments seem to point towards.
Look, I’m a bit capitalist about this. Such games are not going to be overwhelmingly popular unless they actually work. Robert Yang doesn’t think the Dear Esther template can be cloned that much- it isn’t the start of a new genre, it’s just an interesting experiment.
That sounds very similar to an idea I had for a TV show a few years ago. Ha! Amazing.
My point was purely hypothetical because the chances of a computer generating something that I’d consider thematically consistent enough to derive anything of any worth from is probably very slim indeed. Having said this, given the ‘random shit dressed up as art’ link you provided, and your thoughts on David Lynch’s work, certain pieces of ‘art’ have little in the way of obvious thematic consistency or artistic intention anyway so… yeah, blah blah whatever I don’t know. All I’m saying is that ultimately how something is created shouldn’t really matter if the results are the same and provactive or enjoyable in some way, whether you’re fooled or not.
By the way, that article by Andrew Graham-Dixon was interesting, thanks for the link.
Gregg, I’m probably going to write about this later. Well, much later. I remember your discussion on Tap about work standing alone vs. intent. I feel this connects with a number of other scenarios. Thoughts are WIP so don’t want to show my hand too early. I don’t agree/disagree with you – it’s complicated.
Heh, yeah tell me about it! 🙂
A bit of a rebel and in the company of troublemakers. I suspect a pattern is emerging.
Great interview – seems like Dr Fernandez was much more enthusiastic in her responses than Dr Bogost! Very interesting to read her perspectives on the various aspects of her work so thoroughly fleshed out. I’m particularly interested in playing Symon and will have to set aside some time to investigate that soon.