Monday, March 10, 2008

Redrum in the Rue Morgue

As an anthropologist, this panel was, of course, intriguing to me. Expanding beyond your own comfortable perceptions of the world is a difficult thing for most people, and the software and gaming industry are running into the usual (and some unusual) problems as they expand beyond the U.S.:

'Redrum in the Rue Morgue': Collaboration in International Communities

Do Italian citizens of Second Life stand physically closer than Canadians? Are Portuguese more prone to blog than French? Why was a game with ugly 'bad guys" a commercial failure in Korea and a success in the US? As a programmer/computer game designer/web 2.0 guru working with or for international communities and markets, you must have wondered about the cultural models you use and how dangerously close to stereotypes they sometimes feel. If you have read thru this, we assume you belong to the group of those for whom this is a concern. In other words, for you culture does matter. Whether you work with markets to target, video games to market, developing countries to help or academic communities to engage, you must have thought about that thin line between cultural models and stereotypes. Can you design/develop environments and spaces, whether they are virtual or real, that are global without being amorphous? Whether you are interested in designing the web presence for such-and-such company introducing products in international markets or whether you are interested in understanding what makes collaboration thrive or die in international communities, this is the panel where you should be.


The moderator, Ana Boa-Ventura, a professor at the University of Texas at Austin, is Portuguese. She talked about trying to find appropriate clip art for a Portuguese site on women's health, breast feeding and smoking. She had a difficult time finding images that would accurately represent a white collar working class Portuguese. Instead, what she found was either a stereotypical and bucolic images of what others thought the Portuguese were like, or images of women who were blond and white. She'd found she had to search for images of Latinas to get something that wouldn't look out of place on her site.

The other panelists gave very brief presentations which, quite honestly, could all be valuable full-length panels on their own:

Aaron Marcus does a lot of work on usability for companies like eBay and AOL. His company spent some time looking at cultural differences in the use of the Federal Expresses user interface and have done some work on creating multi-cultural and multi-racial images for travel sites. He's the editor of User Experience.

His company found important differences in the use metaphors, mental models, navigation, interaction, and appearance for web sites and online applications. He pointed to the work of
Geert Hofstede who's study in the 70's and 80's determined five "culture dimensions". Marcus's company did their own study, and came up with 29 "dimensions of cultures" and a ranking for the most important dimensions.

His conclusion was that Web 2.0 is "reeking" with cultural differences. Marcus wants to develop a database to help web developers know what is important to keep in mind when designing. He sees culture-centered design making progress and thinks it has to.

I have to admit I was surprised that there was no mention of anthropologists or ethnography, and therefore a bit skeptical of these "dimensions of culture". But without knowing more about them or how these studies were done, I can't really judge if they result in effect guidelines for web developers trying to do interactive design for cultures other then their own. My suspicion is that they are of limited help, since class, race, gender, and a whole host of other factors affect cultural experience. It seems to me you'd need to know your audience in a much more granular way (Portuguese working-class women, ages 20-35 who are pregnant) and work from there first, which will always require some time getting to know your target audience's habits and perceptions.

The next panelist was Prentiss Riddle a senior systems analyst at the IC2 Institute at the University of Texas at Austin. He considered the cultural models harmful and saw problems with national cultural models. he talked a little bit about the granularity idea. He also pointed out a truism in web design - "Don't pattern your IA after the org chart!" and semi-joked about how nationalities are just a kind of org chart. He offered an alternative he called the Sixth Sense Theory, "I see subcultures." (Cute.) He talked about how this was perhaps a little more scary approach for designers to take because sub cultures can be hard to pin down. He also mentioned class divides, and that cultural models are really hard to operationalize. (i.e. Does it really mean that we should use a blog versus a wiki because of the Portuguese national model!?)

He also added that basic access trumps culture - electricity, hardware, bandwidth, literacy, and usability are all issues that have an impact long before cultural perceptions of the design. And then there's individual personality. Adoption of a technology depends on an individual in a certain context (for example, early adopters.) He advised that it's best to follow the usual best practices, i.e. talk to your users, be agile, and be prepared to iterate.

To me, Prentiss made a lot more sense, both as a webmaster and as an anthropologist.

Steve Howard, senior manager of 50x15 Initiative was next. His organization aims to have 50% of the world's population connected to the Internet through affordable Internet access and computing capabilities by 2015. He had some fabulous photos of thier efforts around the world, including an Internet cybercafe tucked inside a shanty town in Nairobi. He pointed out that
emphasis may still need to be placed in web 0.2 rather than 2.0 (Yes! Thank you! Not everyone in the world is an early adopter with high speed connections and the latest technology! Even my parents don't the kind of access needed for some of the highly touted Web 2.0 applications.)

He also mentioned a need to build expertise in these. For example, their organization uses tech savvy youth to train Sao Paulo seniors. he made about about how connectivity must sustainable. he showed a photo of some folks in Uganda, standing in front of their satellite. Unfortunately, it's pointing to ground, probably knocked out of alignment by winds, tree branches, children playing, etc. They were proud to show them the link to the world, even though it was no longer working. Through other photos he showed how they needed reliable structures to house the equipment (a facility in Nepal), and usable equipment. He showed a photo of children in Malawi looking at one of the XO devices, part of the One Laptop Per Child program, as an example of a device and said that it was imporant to remember that it'll take many types of devices to get the world connected.

Afterwards he showed off the XO device, which answered some questions I had about literacy rates as a barrier to using the technology.


A frind of mine and former co-worker, Sean McMains, got one of the XO's in the original "Buy One, Give One" effort. You can check out his thoughts on the XO on his blog.

Lastly we heard from Tim Langdell, a professor at National University and founder of Edge Games. He noted that game designers are really having to look harder and harder at cultural differences to design games that will be successful. Development teams themselves may be in different countries and the gamers are playing together across cultural boundaries.

He stated that video games have a push-pull effect on world cultures (i.e. American English has taken a dominant role over British English). Using World of Warcraft as an example, he pointed out the way developers had setup the choices gamers could make in picking a character. WoW characters are divided into two groups: "Alliance" versus "Horde". The "races" in the Alliance group were Dwarves, Humans, Elves, etc. and all styled to look somewhat attractive. The Horde included groups like Orcs and were styled to be somewhat ugly. In the U.S. gamers pretty much chose to play Alliance or Horde with a 50/50 split, keeping game play balanced and interesting.

Then Blizzard looked to launch the game in Korea. Apparently it bombed, so some research was conducted. (After launching the game. Tsk, tsk.) It turns out that Korean game players didn't like playing ugly looking characters. (There's a much higher affinity for "cuteness" in a lot of Asian culture.) Blizzard had to add "pretty" characters and "ugly" chracters to both divisions in order to get balanced play.

I found this interested, since I'd recently had a lunch discussion about some of these same issues with an archaeologist who also consulted for a game company.

They opened it up to Q&A afterward. Langdell made mention of how the American and Japanese design cultures tend to dominate the world, and neither really considers where it's going to play in the design process. He pointed out that hackers often would break in to replace the language and replace the images with ones more culturally appropriate to them. The reaction from gaming companies was both annoyance and "hey cool". He laughingly suggested that game companies should use hackers to help design. (Honestly, this might be a very good idea. At least, it's very web 2.0 appoach, and very collaborative.)

Marcus pointed out that peripheral stuff will need to be looked at too. While many are concentrating on addressing cultural issues in the technology or applications, we also need to be looking at things like the documentation and help files.

I asked a question about whether any of the gaming industry has considered doing what IBM did, hiring anthropologists to work with their software development teams. Langdell answered, that gaming companies haven't really considered this, and are just stuck in translating games they've already developed. (Play is fundamental to cultural groups. Games really ought to be developed from the ground up, not just relabeled and translated.)

Also mentioned was the reaction of Orkuts when the U.S. based developers went nuts about the heavy foreign use of the product.

I never did quite figure out the reference in the title of this panel.

Catch the podcast here: 'Redrum in the Rue Morgue': Collaboration in International Communities