The Wikipedia Measure

Wiki

Most measures of “internet development” put a numerical figure on basic infrastructure and use: broadband rates, number of users, number of blogs, etc. This data, however, can rarely capture a sense of how a population is using the net and the larger “internet culture” — i.e., does the internet have an impact within the broader culture and/or are users actively invested in the net as a primary cultural sphere? Most comparisons and opinions on these questions are bound to be based on subjective judgment, but are there not objective ways to measure the abstract idea of “national participation” across the globe?

Wikipedia may provide the perfect field for comparison. Almost every major language has its own Wikipedia, and in most developed countries, the site is one of the most central and widely-accessed reference sources. And since users, rather than a specific company or central organization, are responsible for adding entries, the number of pages in a certain language should reflect an overall interest in a population’s desire to contribute to the internet.

In order to compare across the board, we first need to find a ratio for “number of Wikipedia pages in language / speakers of language.” For the number of pages, I used the figures given on the front page of http://wikipedia.org. For the speakers of each language, I used the highest estimate of native speakers on each language’s Wikipedia page, under the loose assumption that most editors and writers are native speakers. To make a more-easily viewed ratio number, I then multiplied the product of the pages/language division by 1,000.

The results:

 Rank  Language  Pages  Total Speakers  Ratio
 1  Dutch  521,000  22,000,000 23.68
 2  Polish  582,000  52,700,000 11.04
 3  German  870,000  105,000,000 8.29
 4  English  2,760,000  400,000,000 7.77
 5  Italian  544,000  70,000,000 6.90
 6  French  770,000  175,000,000 4.4
 7  Japanese  565,000  130,000,000 4.37
 8  Portuguese  461,000  200,000,000 2.30
 9  Spanish  447,000  322,000,000 1.39

As we see, European languages like Dutch, Polish, and German show the highest proportional participation in Wikipedia content. English is relatively high, although perhaps this number is skewed by the high usage of the site by non-native speakers. Furthermore, Wikipedia began in English, meaning the English version of the site has had more time to grow and expand than other languages. Some of the difference seen above is clearly socioeconomic: languages like Spanish clearly have a lot of speakers below the poverty level, who may lack access to computers. Europeans, however, still seem to have an extremely high rate of active Wikipedia participation — perhaps suggesting a strong commitment to Wikipedia’s goals.

I initially explored this idea for a measure in June 2008, and now comparing the number of pages in each language at that point in time with those as of late Feb 2009, Spanish has the highest growth rate at a 19.84% — which means there is a growing effort to make up for the generally low ratio. Despite high page numbers, Dutch still is growing at a relatively fast rate of 16.29%. Italian is similar at 16.49%. Japanese, however, is the slowest of the pack at 13.23%.

This, of course, brings us to Japan, which has a relatively low ratio of pages/speakers and low growth rate. Japan is the most illuminating for this measure, as essentially all statements about the “Japanese language sphere” are basically statements about the Japanese nation. I personally believe the low ratio reflects a certain disinterest in internet participation, which can be seen in other aspects of Japanese internet development. This data, although not particularly in-depth or scientific, suggests that there is a relatively slow embrace of the Web 2.0 concept — including the somewhat ideological aspect that everyone, not just elites, should contribute to the collection of information and knowledge. There is a case to be made that this “information wants to be free” philosophical disposition has no real precedents in Japanese institutional ethics or philosophical values, and regardless, has yet to reach joushiki “common sense” level in Japan. Wikipedia is Alexa’s #10 most-viewed site in Japan, so many Japanese are using the site as an easy and convenient reference, but numerically-speaking, they have much less interest in contributing than most major European language groups in the same socioeconomic order. Are the low levels of participation a product of laziness, disinterest, or a belief that collecting knowledge should be left to elites? Is the prominence of mobile phone access to the net, in particular, fundamentally limiting?

Obviously, this measure is not the ultimate means to gauge abstract political feelings about the promise of the internet, but in the case of Japan, I believe this Wikipedia benchmark gives some indication about the lack of net participation, not usually visible in more general statistics about web usage.

W. David MARX
March 4, 2009

W. David Marx (Marxy) — Tokyo-based writer and musician — is the founder and chief editor of Néojaponisme.

44 Responses

  1. MattAlt Says:

    I wonder if this isn’t another case, similar to blogs, of the depth far exceeding the breadth of usage. (Remember the 2007 incident involving Agriculture Ministry employees using government computers to update the Gundam entries.)

    http://news.bbc.co.uk/2/hi/asia-pacific/7029685.stm

  2. W. David MARX Says:

    Wouldn’t the number of wikipedia pages increase from more depth as well as more breadth? If you have 100 guys writing 100 pages about obscure otaku topics (which is what Wikipedia-J is like generally anyway), is that going to make less than 1000 guys doing 10 pages on lots of different topics?

  3. M-Bone Says:

    While I don’t disagree with the result, I’m not sure how you could prove this –

    “I used the highest estimate of native speakers on each language’s Wikipedia page, under the loose assumption that most editors and writers are native speakers.”

    We can assume that for Dutch, but for English? “Most” maybe. But the “other” part may add hundreds of millions of potential users. The “History of Japan” article seems to be in the middle of a war between Japanese and Korean netizens. Surprise, surprise.

    I just looked up the wikipedia traffic breakdown on Alexa and it seems that Japan provides the second largest number of hits to wikipedia.org so its not like the resources are not being used. I wonder if the big net discussion forums like 2ch (still holding its own at traffic rank 15 for Japan) have simply pushed people into another form of participatory online experience?

    Shouldn’t we also note that the Wikipedia editing experience for English speakers is mostly anonymous (at least for the few articles that I have looked at changes for) with sock puppets galore and people not providing “public identities”?

  4. W. David MARX Says:

    We can assume that for Dutch, but for English?

    Yes, this may be the X factor. But I seriously doubt for languages like Polish, Dutch, and Japanese that you have non-native speakers doing a significant part of the editing. And if you compare just these three, the number of Japanese pages is still very, very low.

    The really good research project on this would weigh demographics, socioeconomics, and “wired”-ness. But I wonder, is Holland really a much younger or richer country than Japan?

  5. W. David MARX Says:

    I just looked up the wikipedia traffic breakdown on Alexa and it seems that Japan provides the second largest number of hits to wikipedia.org so its not like the resources are not being used.

    I think my point was that Wikipedia is being used as a reference, but writers are not participating in sheer terms of volume of pages added.

  6. M-Bone Says:

    “But I seriously doubt for languages like Polish, Dutch”

    Could argue, however, that for languages like Polish and Dutch, there just isn’t nearly as much public space discussion and popular culture as in English and Japanese (two almost uniquely huge markets) to go through so people end up gravitating toward something like Wikipedia instead. I wonder how a smaller country known for its high rates of newspaper readership – Norway – ends up faring? It could be that vitality in this area puts downward pressure on Wikipedia editing.

    “I think my point was that Wikipedia is being used as a reference”

    I got that, but I was still surprised to see Japan at number 2.

    Another good question – wired, pretty wealthy, socially similar to Japan in some ways, and perpetually pissed off – where is Korea? 91,000 articles for about 50,000,000 South Koreans (this is one for which “native speakers” is not a good one to use). This is under half of Japan’s ratio (and does not count things like the huge Korean diaspora and those two guys in the north who edit porn star profiles). I think that there are so many potential factors behind differences that comparisons between how Wikipedia fits into other enviornments are not that useful.

  7. W. David MARX Says:

    The Korean counterexample is interesting since Korea is so “wired.” Is there another site like this in Korea that competes?

  8. M-Bone Says:

    My first instinct is that, like in Japan, the huge discussion sites draw away a lot of the participatory web types.

    If much of the effort that gets put into 2ch (I mean anime commentary and stuff, not just hating Koreans) was focused on Wikipedia, we’d probably see a huge change.

  9. Chris Says:

    One particular factor with Spanish is that there was some controversy with Spanish Wikipedia for a while (e.g. see http://www.spiritus-temporis.com/spanish-wikipedia/), that may have slowed its growth down.

  10. Mulboyne Says:

    A lot of Dutch wiki contributors add pages which are basically translations of existing English and German wiki entries.

    I was just looking at the English wiki entry for Shiina Ringo the other day and it reads very much like a J -> E translation, probably by a native Japanese speaker. If that’s right, then it’s an example of a contribution which would be wrongly categorized by your analysis.

    The English-language wikipedia has a large and wealthy audience so there may be an incentive for people from non-English speaking countries to contribute to it in order to influence how particular topics are discussed in the English-language world.

  11. Karamoon Says:

    When writing about “the Internet”, Internet should be written with a capital “I” to differentiate the Internet from the millions of other internets that exist.

    Japanese people do not make significant contributions to the Web because Japanese people don’t, in general, make contributions to anything, neither in Japan nor elsewhere. It is primarily a society of consumption. To give just one example, look at how few Japanese people speak at international conferences. Compounding this is the fact that Japanese people are very hesitant to take ownership of projects or even parts of projects.

    I do not, for one second, suggest that this is necessarily a bad thing…

  12. LS Says:

    I hate to point out the obvious, but couldn’t it be that different cultures just do different things online? Like, in the case of Korea, playing MMORPGs rather than editing articles. Why aren’t we doing this in terms of number of WoW accounts? I’d say that collaborating on a raid counts as “participation in the Internet.” Isn’t it just your bias toward certain types of participation over others?

    Also: do we capitalize “Arnold Schwarzenegger” to differentiate the Governator from all the other arnold schwarzeneggers?

  13. W. David MARX Says:

    I hate to point out the obvious, but couldn’t it be that different cultures just do different things online?

    Absolutely. I am using Wikipedia as one way to show that: proportionally less people are contributing to Wikipedia (or contributing less). Whether you think that is bad or good is up to you.

  14. LS Says:

    Wikipedia may provide the perfect field for comparison. Almost every major language has its own Wikipedia, and in most developed countries, the site is one of the most central and widely-accessed reference sources. And since users, rather than a specific company or central organization, are responsible for adding entries, the number of pages in a certain language should reflect an overall interest in a population’s desire to contribute to the internet.

    Sure sounded to me like you were trying to set up a universal yardstick of Internet participation.

    It’s no mystery that Japan, with an Internet that’s mostly accessed through mobile phones, might take on different characteristics.

  15. Bobbin Says:

    “Wikipedia may provide the perfect field for comparison.”

    I’d agree with LS. It seems like Wikipedia is good for measuring the extent to which people want to contribute to … Wikipedia. And certainly Wikipedia carries with it a bunch of assumptions about what form the “desire to contribute” should take.

    A measurement that is composed of a few different indices — including blogs — would probably be more revealing. Right now, for example, we’re all reading this blog entry and having a discussion. We might do this a few times a day, each. Do we contribute to Wikipedia with the same frequency? I’d tend to doubt it.

    How about blogs, then? How do we measure them as well?

  16. Karamoon Says:

    @LS: The Internet refers to the global network of networks, on top of which services such as the Web, email and IRC run. “The internet” means any internetwork which the speaker and listener(or writer and reader) are familiar with, excluding the Internet.

    Regarding Schwarzenegger, the less he is referred to, the better.

  17. Adamu Says:

    I was in a McDonald’s the other day and saw an couple in their early 20s using their laptop. They were looking through J-Wiki to figure out whether Antarctica was a country or not. So at least in that instance, Wikipedia was fulfilling its role of making it obsolete to ask people factual questions.

  18. LS Says:

    @Karamon —

    My point was, the Internet is a proper noun by virtue of being a named project, not as a means of distinguishing it per se.

  19. Karamoon Says:

    @LS: A fair point. Rereading my comments I see that what I meant was not that the use of capitalization is aimed at distinguishing an internet from “The Net” but that the lack of capitalization leads to confusion. As a fan of Pocky and ninja, I often find myself reading and writing about internetworking with TCP/IP, and “The Net” in the same document.

  20. Matt TREYVAUD Says:

    I have to agree that this measure alone tells us little. Here are my issues:
    – Number of speakers is a bad independent variable. Keitai effect, etc. Number of person-hours at an actual PC is what’s required.
    – Number of articles is just one dimension–what about length of articles, number/frequency of edits, etc.? Maybe Japanese folks are Wikipedia-editing fiends but don’t like to create new articles entirely
    – Cultural issues. We’ve all read about how much influence editors have at Wikipedia, and how most of the work is done by a small fraction of contributors. What is JW’s culture like? I don’t know. There could be other cultural issues too, e.g. maybe there are more articles in European languages because translating them from English is so much easier than translating them J2E due to the relative closeness of the languages.

    I’m just not convinced that you’ve shown anything here that can be extrapolated beyond Wikipedia itself.

  21. Birdseed Says:

    I think it might be worthwhile to try the hypothesis on smaller languages as well, and see if it produces any discrepancies. Here’s the full list of wikipedias:

    http://meta.wikimedia.org/wiki/List_of_Wikipedias

  22. Ark Says:

    There is an important flaw in your statistical analysis. Simply looking at the number of pages per Wikipedia is not, by itself, a sufficient indicator of what you are trying to look at. For example, it will not show you if one language’s wikipedia has a tendency to focus on making many edits to a page to increase it’s factuality and general quality, or if another wikipedia tends to have more stubs and slapped-up articles of poorer quality. Trying looking into the “tell us about your wikipedia” sections, where you can see various people talking about the strengths and weakness of their various wikipedias, such as the ones I mentioned above.

  23. Matt TREYVAUD Says:

    Also, to expand on M-Bone’s point:

    I wonder if the big net discussion forums like 2ch (still holding its own at traffic rank 15 for Japan) have simply pushed people into another form of participatory online experience?

    If you look at Wikipedia as a place where nerds come to argue (rather than a place where the sum total of human knowledge etc.), 2ch starts to look like a huge factor affecting Wikipedia’s success in Japan. There are fundamental differences between the two (e.g. anonymity of 2ch vs nonymity of influential Wikipedia editors) but the fact that there already was a huge, centralized place for nerds to argue had to have some non-negligible effect.

  24. Freakman Says:

    One of the most surprising things about Japan’s Wikipedia is the almost total absence of sources. And there’s no one complaining about it (i’ve never seen the warning “this article does not cite its sources”) I’m not saying Wikipedia ought to have sources for everything, but I wonder if it’s representative of something.

  25. Ian Duncan Says:

    The comparison with France seems worth noting. Perhaps the French Wikipedia is growing faster but there is still a very similar ratio. Would you make the same argument that France is an intellectually elite country? I probably wouldn’t.

    On a different note: if you read English Wikipedia on Japanese topics you get the feeling quite quickly that a lot of the editing has been done by Japanese people.

  26. M-Bone Says:

    “look at how few Japanese people speak at international conferences”

    Look at – patent applications in the US, articles published in English-language academic journals, books published, DVDs released in English by foreign language (Japanese is second after French, not counting anime), size of the do-it-yourself manga market, poetry contests and magazine submissions, etc.

  27. W. David MARX Says:

    Number of speakers is a bad independent variable. Keitai effect, etc. Number of person-hours at an actual PC is what’s required.

    There are 100 million mobile phones in Japan, apparently. I would say that you have a very high percentage of people wanting to access the internet accessing it.

    I think we are all on to good reasons why Wikipedia Japan is lagging, which is part of what I am after. If, for example, Wikipedia Japan had the highest ratio of any country, we would also be saying it means something, right? When Technorati talked about “Japan has the highest number of blogs around the world” — later proven false due to the spam blog problem — everyone came up with some kind of big narrative for the reason why.

    I don’t think this Wikipedia measure is the end of a conversation; it’s clearly a start. But the site is widely used in Japan and yet, for whatever reason, has a proportionally small number of pages. Are there not other parts of the Japanese internet experience that reflect this? Fewer “up-start” blogs trying to take a central position in the public media, for one

  28. Matt TREYVAUD Says:

    I would say that you have a very high percentage of people wanting to access the internet accessing it.

    I’m not sure quite what you mean here… My point is that since most people can do 90% of their internet community stuff via their keitai (email, Mixi, Ameblo, etc.), they spend less time sitting down at actual PCs, which means less opportunity to edit Wikipedia.

    If, for example, Wikipedia Japan had the highest ratio of any country, we would also be saying it means something, right? When Technorati talked about “Japan has the highest number of blogs around the world” — later proven false due to the spam blog problem — everyone came up with some kind of big narrative for the reason why.

    Well, yeah, so everyone was wrong then, and it would be wrong to jump to conclusions in the same way now.

    You say clearly that your explanation for the data is what you “believe” to be true, so I’m not accusing you of chicanery or anything. Nor am I saying that your ideas about online participation in Japan are necessarily incorrect. But I just don’t think the data presented here is robust enough enough to draw any broader social conclusions from. In fact, it seems to me that you could have drawn similar conclusions from a Japanese Wikipedia with more articles per naked speaker (maybe along the lines of “See? Japanese users gravitate disproportionately to sites where they can stay pseud-/anonymous but still contribute meaningfully”).

  29. Matt TREYVAUD Says:

    naked speaker

    Sorry, I must have been thinking about some other website.

  30. M-Bone Says:

    I think that we also need to problematize the idea that editing wikipedia is resistance to established hierarchies of knowledge.

    First, I’m sure that many or most here will agree that the “nerd” factor that I raised and Matt expanded on is very important. The average article on the English wikipedia is about old commercials for Tang or Al’s third girl friend on “Home Improvement” or something like that.

    Second, the serious articles in Wikipedia, to me at least, seem to be all about replicating existing knowledge hierarchies. Borderline plagiarising secondary academic sources is fetishized on wikipedia and the result is that a lot of the serious articles are a mishmash of different stuff slapped together with no eye on quality or originality. Shockingly, I have seen quite a few debates where original sources and documents are kicked out – they demand books from famous publishers that affirm the right of the wikipedia contributors to actually do some creative, interpretative work. Wikipedia could solve this problem by allowing different versions of articles – ie. actually admitting that there are irreconcilable differences of opinion on how some things can be narrated with both versions more or less equally correct. The idea of Wikipeida, however, seems to be that there has to be one vision that represents “human knowledge” where everything has to be based on… stuff ripped often at random from academic books.

    In the end, serious Wikipedia pieces almost always remind me of C+ first year undergrad papers.

    Don’t want to hijack, but have you guys heard about this yet?
    http://www.imdb.com/title/tt0991175/
    Uggghhhh

  31. Graham Says:

    From an Internet research perspective I can think of a bunch of problems here, but perhaps the biggest is the focus on one website, but here are two:

    1. The Chinese wikipedia is not Wikipedia, it’s Baidu Baike 百度百科. If Chinese numbers were included above, they would look something like this:
    service——articles——% of 1.18 B “Chinese” speakers
    Wikipedia——237,000——0.02 %
    Baidu Baike——1.48 M——0.13 %

    Not a high percentage, but this shows that Wikipedia use doesn’t work as a proxy even for wiki use.

    2. Speakers of a language versus users of the Internet. Internet penetration is around 70+ percent in one form or another in Japan, the United States, South Korea, and a few others. For Japanese, a similar number of speakers of the language will be online. For Korean, there’s the problem of the difference between north and south. For English, penetration is different in different countries.

    Back to China, just over one-fifth of PRC residents are online, according to the government’s numbers. Once you overcome the problem of selecting based on one website, you’d need to clarify whether you’re asking about how people from a country or a linguistic sphere use the internet, or how many. It’s important to remember you’re commenting on the ones who are online, and, realistically, a small minority of them.

  32. Aceface Says:

    I so want to see that “Speed Tribe”!
    It’ll complete the trilogy of misrepresented homeage to Japanese highway stars by Chinese film makers along with “Initial D” and”The fast and the furious,Tokyo Drift”….

  33. Connor Says:

    Aceface I am with you. We absolutely need to watch that shit.

    “Marauding motorcycle gangs rule the streets of Tokyo. When Tats – the young and cavalier leader of the Midnight Angels – botches the delivery of a mysterious package, consequences mount and a high-speed and adrenaline-fueled adventure ensues. With the help of a street savvy hostess, Tats drops in on a gaggle of Tokyo’s most unlikely and colorful urban inhabitants, putting clues together and avoiding the reprisals of rival gangs as he races to make amends for his mistake and restore balance to his world. Based on the best-selling book by Karl Taro Greenfeld.”

    They forgot Seanbaby’s “Pink-haired rollerblading hackers” but otherwise they are basically going all out. But WILL THEY feel the need to include more awful Teriyaki Boyz songs? I wonder if they know? How we live in To-kee-oh?

    Could it not at least have been “how we do?”

    Back to the discussion at hand- with regards to M-Bone’s point about sock puppets, I think it’d be useful if we could get ahold of the number of unique IPs generating connections to the Edit side of wikipedia from each country. And also, I don’t know anything about this really but- aren’t there sizable communities of Nikkei people in Hawaii and Brazil and Peru and some other places? If their computer use was trackable, wouldn’t that put to the test notions of Confucian-derived ethics/mores standing as a sort of barrier to bottom-up Japanese net participation?

  34. W. David MARX Says:

    Second, the serious articles in Wikipedia, to me at least, seem to be all about replicating existing knowledge hierarchies.

    Yeah, but I think Wikipedia reflects a lot of why the internet succeeds in the U.S. — because people are allowed to have their own opinion elevated to the same level as “experts.” This is mostly bad since no one knows anything, but there is no holding back like, “Me? Edit Wikipedia? I am no PhD, sir!”

  35. Adamu Says:

    Unlike David, I am just not seeing this “disinterest in Internet participation” – people definitely want to participate in the Internet here, but they may find themselves stifled in important ways.

    Thing is, I feel like the Japanese media companies have been more successful than others in slowing the pace of Internet growth through later technological adoption (magazine articles began appearing about web 2.0 something like six months after it was a quite common term in the US) and anti-Internet propaganda by a very centralized national media.

    I also have my doubts about this “Wiki gap” argument. Since English Wikipedia is the Mother Brain of all wikis, it doesn’t make sense to limit the English speaking world to the traditional 400 million. This would necessarily lower the ratio, perhaps even below the one for Japan, not to mention the other problems mentioned.

    J-Wiki is already one of the top wiki sites in terms of article numbers and already fulfills a role as a source for a whole range of topics important to Japanese internet users. So it might not compare to the universal language, but already the Japanese Wiki is pretty robust. To the extent there might be a lag in Wiki development, I would see it more as a function of top-down Luddism from the big companies than bottom-up “disinterest.” And it’s ironic that I am saying this since it’s usually Marxy who’s reminding people not to confuse hierarchical trends with grassroots movements. And if people who don’t know what they are talking about avoid editing Wikipedia, isn’t that a good thing?

    Right now I think the more advanced functions of the Japanese internet (blogging, etc., things that go beyond simple email or browsing that are already widespread) are in the process of transitioning from a space only for tech-savvy nerds into a space for everyone, with the big media/tech companies being dragged along kicking and screaming. And despite better internet connectivity, this transition has come later than in the US. While traditional deference to authority and blah blah blah might be a factor, isn’t it true that the Japanese internet mucky-mucks wait for the innovations to come from the English-language Internet before they are adopted in Japan? Wiki Japan itself didn’t get started until a translated article on Wikipedia appeared in Wired Magazine Japan. We’ve seen this again and again, with important exceptions, starting with e-commerce during the dot-com boom and continuing into web 2.0. This is possible because of the language barrier that isn’t broken by the new Internet age, but it’s harder to control – youtube got huge in Japan after some simple Japanese-language interface software became available, before the US youtube came out with an official Japanese language site.

    I want to emphasize that J-Wiki is already a pretty amazing resource for topics specific to Japanese society – ranging from suicide to electronics regulation to English teaching. And of course, it’s second to none for information on Gundam and Japanese trains.

    Also, on the Japanese internet it is common for people to form independent wikis for certain topics. This practice is also common in English but worth mentioning here as well. For instance, when there was a danger that inept bureaucrats were about to accidentally ban used electronics back in 2006, a very robust wiki was formed by angry netizens to keep tabs on the situation. Similarly, there is a wiki dedicated to documenting the various oddities and hypocrisies of the mysterious Kikko blog. And these are no PHDs, though maybe they are directed by respected higher-ups such as Sakamoto Ryuichi in the electronics case and right-wing alpha bloggers in some of the more prominent flame attack cases.

    So demand is definitely there, even in this narrow area of wiki-ing. Other areas such as closed Mixi diaries are also huge.

    The problem is that there is a critical lack of understanding by the elites (at least the ones at the helm) that the potential for the internet is beneficial for society, and this has hindered the development of the Japanese internet in two ways (1) a near-complete reliance on foreign innovations that must be translated into the mother language, which takes time; and (2) a failure by the government to establish clear rules of the road and punish troublemakers (2ch!!!!), leaving many of the challenges posed by the internet such as copyright enforcement, harassment, etc., unresolved.

    And this lack of imagination can be helpful to the big companies that have foreign models to draw on, but it can have dire consequences for the clear medium-term losers in the Internet age, such as newspapers. Institutions like the Mainichi are scrambling to find a survival strategy as more than a decade of sitting on their laurels catches up to them and their counterparts in the US are faring no better.

    Anyway, while we are looking for circumstantial evidence for how Japanese people use the Internet, I thought I would pass this along –

    For Yahoo Japan, the country’s biggest portal site, the percentage of page views from mobile phones has steadily increased from 8.29% in April 2008 to 12.76% in January 2009. It’s still a pretty low number.

    People might be using their phones more to post to 2ch/@wiki, update their mixi diaries, or access game sites like Mobage, but I think it’s telling that even a major portal like Yahoo has such a low keitai access rate.

    http://i.yimg.jp/images/docs/ir/monthly/2008/monthly0210.pdf

  36. W. David MARX Says:

    Thanks for that comment. I think traditional market players’ refusal to offer content on the net added to a general lack of IT entrepreneurship in Japan has been devastating to internet diffusion and progress.

    I think the English Wikipedia is the most “exceptional” and the ratio is not a good number. I am more interested in why Japanese is under Polish and Dutch. I don’t just buy that “the Dutch translate more from German.” They have to have some reason to be doing it. Language barrier certainly does not help the J-Wiki…

    I will concede that the main issue is probably not “philosophical” but the fact that the “wiredness” of Japan does not capture how little the internet has penetrated into people’s lives. In the US, Web 2.0 was that step where people went from passive to active usage. I don’t think you can say that’s happened in Japan yet to any huge degree.

    The net was “interesting” in the US circa 1997 but it’s a hell of a lot more meaningful now.

  37. M-Bone Says:

    “This is mostly bad since no one knows anything”

    You could argue that Wikipedia is not the place to do this, but as it is set up now, it actually prevents people from developing their own ideas or doing their own original research by demanding that they cite “experts” like, for example, a certain Speed Tribes author.

    So people do jump right in, but they aren’t jumping in in a way that fits with Wikipedia’s lofty PR.

    Ironically, while academics are now giving more cred to blogs and online sources, Wikipedia is one of the major knowledge spaces out there that consistently ranks blogs of all kinds as second rate compared to published stuff (although this depends on the community that has sprung up around a given article). I think that this part of the net is in the process of devouring its own tail.

  38. dotdash Says:

    It’s a pretty mundane explanation, but don’t you think working hours might have something to do with it? Korea has by far the longest working hours in the developed world whereas the Germans and Dutch have among the lowest. Japan is still pretty high up in the rankings and that’s before you even take into account all the obligatory after-work drinking sessions. Perhaps people just prefer to spend their brief snatches of free time doing other things than writing Wikipedia articles.

  39. Mulboyne Says:

    The Dutch Wikipedia has a great number of entries where the primary source is clearly the English Wikipedia entry. I strongly recommend you read a few entries at random to confirm that. German language entries are translated less frequently but are probably the next most popular source. They seem to be more often incorporated into an original Dutch entry rather than lifted wholesale. If you don’t trust your Dutch to make a comparison, just look at the notes. On the Dutch Wiki, the sources are also supposed to be largely in Dutch so most of the references from the English original are simply left out while the content is retained. For instance, the English Wiki entry on Hitler cites over 300 references, the German entry cites 78 while the Dutch entry cites only 8.

    That’s not to dismiss the Dutch Wikipedia as a simple knock-off of the English version. It’s much easier to translate or summarize an existing article than to create one from scratch. However, I’m guessing that once an article is up, it can develop a life of it own, bringing in contributors who are happy to add a sentence or two but who wouldn’t have bothered to write an original piece in the first place.

  40. xee Says:

    one interesting thing about dutch wikipedia is that a high proportion of literate/computersavvy dutch&flemish speakers are functionally bilingual or trilingual, and could probably “make do” reading english and french-language entries, or indeed german-language ones. It seems to me that there might be an element of… not linguistic nationalism, but an active effort to ensure that dutch is well-represented.

    (also i just think of the dutch as being very active internet users, in a late-nineties kind of way? i have this sense of young people in the netherlands being disproportionately fond of irc, for example, and a lot of open source/shareware people are based in the netherlands. less in belgium.)

  41. xee Says:

    also i keep reading the title of the post as ‘the wikipedia massacre’ :(

  42. M-Bone Says:

    Mulboyne – you read Dutch too!? Are you really five different people commenting under one name or something? If not, your ability to comment insightfully on just about any topic is amazing.

  43. wah Says:

    If Nico Video and Pixiv are any indication, I think only otaku use the internet n Japan :S

  44. Petter Says:

    Sweden has 300k pages and 9000k speakers.