Emmanuel M Décarie's weblog on scripting.
Mostly PHP, Radio UserLand, Perl, Applescript and OS X.
RSS/XML Feed | | Picture | Search
Categories: AppleScript | BBEdit | Internet | Javascript | OS X | Perl | PHP | Python | Review | Radio UserLand
Technorati URL search for
Search Technorati for information relating to the URL or keyword of your choice.
... Original post: Emmanuel M Décarie's weblog on scripting. by at Daypop Search - filter replacement ...
Filter Replacement View Technorati URL search


Mon 17 Feb 2003

Review: Clay Shirky, Power Laws, Weblogs, and Inequality

Le duc parut touché que je fusse venu chez eux le jour même de son retour. Mais sa figure se rembrunit quand je lui eus dit que je venais demander à sa femme de s'informer si sa cousine m'avait réellement invité. Je venais d'effleurer une de ces sortes de services que M. et Mme de Guermantes n'aimaient pas rendre. Le duc me dit qu'il était trop tard, que si la princesse ne m'avait pas envoyé d'invitation, il aurait l'air d'en demander une, que déjà ses cousins lui en avait refusé une, une fois, et qu'il ne voulait plus, ni de près, ni de loin, avoir l'air de se mêler de leur listes, de «s'immiscer», (...), car en toute hypothèses les listes de la princesse étaient certainement close. - Proust, Le côté de Guermantes, Laffont, 1987, Paris, p. 465

I have read as others with great interest Clay Shirky paper called Power Laws, Weblogs, and Inequality. I will not discuss the modeling of the data, since it seems valid and show an interesting point of view. My intention in this review is to discuss how much this analysis is missing the whole context in which blogging evolve. This analysis could have been more meaningful if it went further instead of relying only on a data model based on the famous "Power Law" which in itself explain nothings.

I knew that the model was going to be in some sort of "in vitro" analysis when I saw that one of the first reference on which the model is based was coming from Vilfredo Pareto. You see, Pareto is a member of a sociological approach that goes from Wright Mills to Raymond Aron, and that focus a lot on analysing the elites, how they related as a group to each others, evacuating the connection between the elites, as groups of individuals, and the society, as a sociogical, historical and economic system that shape the place of the individuals. In short, the power law doesn't give a damn about context and content.

That lead to gross misreprensations like the following:

In systems where many people are free to choose between many options, a small subset of the whole will get a disproportionate amount of traffic (or attention, or income), even if no members of the system actively work towards such an outcome. This has nothing to do with moral weakness, selling out, or any other psychological explanation. The very act of choosing, spread widely enough and freely enough, creates a power law distribution.

You have to question the above assertion. Are you sure that people are really free to choose? In my opinion, people are more shaped by what Pierre Bourdieu call their "habitus". In short, by habitus, Bourdieu mean that an individual is shaped by its socio-economic position in society and this will create a certain degree of expectation, taste and way to express itself. Its a pessimistic view, akin to Calvinist predestination, where your destinity is somehow written, and where your can't have greater expectation than what your socio-economic position let you expect. It goes really against this "American Dream" fantasy, and its quite easy to understand why this myth still exists in the United States.

So saying that 20% of the population holding 80% of the wealth is due because of the existence of a "power law" explain nothing, and is hardly true regarding history, and developed countries where the elites holding the wealth are sometimes only 1% of the population. If for example, its true that 20% of the population hold 80% of the wealth in the US, there is surely more than a statistical model to explain that, which is surely related to some form of class struggle, systemic condition in the American society, history and so on.

Since this is not a paper presented to an university, or to a scientific publication, I will not pretend that I have checked all the background. In fact, on a personal note, I was a scholar in another life, and I'm no more (but I still know how to do name's dropping). My intuition is that the data collected for figure #1 represent mostly white males, with a university or college degree. How can I say that with any data to back it up? Well, its simple, I've been following the web since 1996, and weblogs since the name was coined, I'm on numerous mailing lists, and usually, the people that I see there fits this profile. I must say that since I'm now a programmer, you have to take the above with a pinch of salt.

What I'm trying to say is quite a tautologie. People will be attracted to a certain type of weblogs because of their habitus, and because of that, they will expect a certain content to be displayed in these weblogs, so this will reenforce in some ways their own excellence (to be able to read one given weblog and even discuss with the weblogger).

So this explain why people are enticed to go to a certain type of weblogs, its because they probably share a lot of the sociological background with the author, and the author give them what they expect (and not much more), so this reassure them that they are what they think they are in term of the position they have in the society.

That's why, such an assertion like the one that follow, feel like something that is totally disconnected from reality, it completely evacuate socio-economic context, and state the problem in term of "freedom of choice" when, in some way, there is no "freedom" because people are conditioned since their childhood to think in a given set of values ("habitus", or if you want something more weberian, "world conception").

Note that this model [the power law] is absolutely mute as to why one blog might be preferred over another. Perhaps some writing is simply better than average (a preference for quality), perhaps people want the recommendations of others (a preference for marketing), perhaps there is value in reading the same blogs as your friends (a preference for "solidarity goods", things best enjoyed by a group). It could be all three, or some other effect entirely, and it could be different for different readers and different writers. What matters is that any tendency towards agreement in diverse and free systems, however small and for whatever reason, can create power law distributions.

And to add more to my point:

The first, of course, is the freedom in the weblog world in general. It costs nothing to launch a weblog, and there is no vetting process, so the threshold for having a weblog is only infinitesimally larger than the threshold for getting online in the first place. .

Now that's interesting. I'm sure for example that "Cosmos" from Witold Gombrowicz will cost around $5 bucks in an American bookstore. But who is going to read such a master piece. Probably, somebody that have a college degree, is interested in European literature and discussing philosophical questions and heard about this genius Polish writer. So the same goes for blogging. Although its true that it cost nothing to do a weblog, you have from the start to be pretty sophisticated (in the general term) to do so. Weblogging is not a mass phenomenon that transcend all classes, weblogging emanate from certain social stratas in the American society, and yeah, guess what, they are part of the elites, but not because of some funny power law, but probably because their parents had a college degree and could afford to send their children to the university.

Now how can we explain the A list?

Third, the stars exist not because of some cliquish preference for one another, but because of the preference of hundreds of others pointing to them. Their popularity is a result of the kind of distributed approval it would be hard to fake.

This make me puzzling. Why some cultural artifact are more popular than others. If you explain this by a "power law" you still miss the whole point, which is content, and how some weblogger are able to articulate content to suit the taste of so many people. The content that result is in some ways, a chamber-effect of their habitus, it cristalyse a common ground, a common expectation that is demanded by their public. Why the Beatles in the sixties? Why the Clash in the 80's? Barthes said a long time ago, in a bold statement, that there were no authors, the author is dead. He wanted to mean by that that the culture is permanent, and author are just passing by, reformulating for their times questions asked by their art. This is why the following statement is in my opinion false.

Inequality occurs in large and unconstrained social systems for the same reasons stop-and-go traffic occurs on busy roads, not because it is anyone's goal, but because it is a reliable property that emerges from the normal functioning of the system. The relatively egalitarian distribution of readers in the early years had nothing to do with the nature of weblogs or webloggers. There just weren't enough blogs to have really unequal distributions. Now there are.

What we need to understand the A list, is (1) to analyse the content they produce as a cultural artifact that take ground in some socio-economic and cultural world, (2) from where is coming the legitimity of the stars webloggers (à la J.-L. Austin) . I think that from there, what Clay have wrote could be much more valid, and offer a more comprehensive explanation of the inequality in the weblog world. I can't simply agree with the conclusion of his article. Society is not a network model, Parson have been dead now from a long time. If there is inequality between webloggers, "la raison en est ailleurs".

comments: 0

8:01PM EST [ /Review | # ]

Wed 12 Feb 2003

Review: Building with Blogs by Doc Searl and David Sifry, Linux Journal, March 2003

I received yesterday the last issue of the Linux Journal which is devoted to Blogs and Lists. I didn't read all the issue, but perusing the content, I find it much interesting compared to the previous issue that I read in the last six month of so (its seems that in this period of time, scripting languages where not covered as it was before).

There is a very good article by Doc Searls and David Sifry called Building with Blogs and its very unfortunate that its not accessible online without a subscription to the Linux Journal. But hey, you can buy it too at your newsstand, its only $5 bucks.

I have find this article quite illuminating, and a very good introduction to weblogs for someone that is not a geek. And I'm quite happy that Radio UserLand is part of what the authors call the "big three", i.e. packages designed from the ground up specifically for blogging; Blogger, Movable Type and Radio UserLand.

I'm also happy to see how Dave Winer is getting recognition for its contribution to the Web (XML-RPC, SOAP, RSS). Apparté: I'm a long time Frontier user (starting with version 4.1), and I learned programming thru Frontier and the UserTalk language, and really lived the transition of Frontier from a Mac scripting system to a multi platforms system that is focusing on publishing on the Web. It was quite a trip!. Another big contributor is Evans Williams of Blogger who created the Blogger API.

There is a very good analysis here on what is a blogging tool and the distinction between different packages that are either built for this task (like the big three) or are, as the authors put it, first generation packages [that] allow you to set up story posting, site membership, comment moderation, topics or categories (...) (like Slascode, Scoop, etc...), or are the second generation tools like PHP-Nuke and PostNuke. Its not clear to me what are the differences between the first and second generation. It seems that the difference lies in the fact that they came later and are based on PHP. These latter packages are labeled by the authors as discussion sites. The authors discuss also blogging tools that comes from content management system, like Zope of from the wiki's galaxy.

The authors offer a nice framework to distinguish between pure blogging systems and the others, as a kind of programming eugenism. A system is considered as a "blogging system" if it can answer yes to these questions (and the authors think that the big three are evidently in this class).

Overall, this list is very good but it present some problems too. For example, Radio UserLand doesn't qualify for point 3 if you are using FTP to upstream your posts. And I'm not sure that Radio could be use as a server for multi users. I might be wrong and anyone can correct this statement by posting a comment.

Point 5 present some problems for me. Although blogrolls have their places, and allowed me to discovers other blogs, I had always the feeling that there are a little bit incestuous, some sort of moebius spiral, or are similar to the French expression "Renvoyer l'ascenseur". But the fact remains that blogrolls are very good for building community of interests.

And last but not least, I think that a comments system could have been mentioned. I'm not sure that its such an important function for blogging tools (it is at least for me), but I think that all the big three offer this functions and more (like trackback for Movable Type).

The article cover also the importance of some services for the blogging community, like Google who is a perfect fit for blogs because the way it rank pages. The authors gave an example of a request with the occurrence "802.11b" which list first WI-FI Newworking News by Adam Engst of TidBITS fame and Glen Fleishman. Here's we see another contradiction. This weblog hosted on Weblogger is using Manila, which is a CMS AND a blogging tool. Does Manila belong in the big three? My guess is yes, and I don't understand why its not included in the big three, because Manila was really build from the ground to have a News page, but maybe at this time (I think 1999 when it was first launched), the word "weblog" was not in great usage.

Other services for the blogging community that are mentioned are Technorati (build by Dave Sirfy, one of the author of the article, which is discussed at length ), the Blogging Ecosystem, Blogdex, DayPop and

And finally, Blosxom from Rael Dornfest is also discussed at length. I'm not sure why its not also part of the big three since it was build from the ground up to be a blogging system, and since it qualify to most of all points beside point 3 (which I don't think Radio UserLand qualify anyway) and point 5 (which I don't think should be in this list too). Sadly, the excellent PHPosxom from Robert Daley is not mentioned in the Blosxom family. This site run on a slightly hacked (by me) PHPosxom version and Radio UserLand, which is for me, the best of both worlds.

So in the end, I think the authors should not talk about the big three, but instead of the big five (Blogger, Movable Type, Radio UserLand, Manila, Blosxom and the Blosxom family). Overall, this is a really good read, and I encourage you to buy this issue of the Linux Journal (I wish the Linux Journal put this article online without restricted access, because it could generate more discussion on its excellent analysis).

Update #1 [February 13 2003]: Mark Paschal note that Radio maybe qualify to point 3 because of the Multi-Author Weblog tool. Here's the way it work:

Everyone in the group writes with their own copy of Radio and publishes their weblog in both HTML and RSS. One of the editors takes responsibility for running the Multi-Author Weblog Tool, which joins all the individual feeds into a single weblog. This person is the webmaster of the multi-authored weblog.

Now, that's an interesting implementation. The only drawback (if you can qualify this as a drawback) is that each author post to their weblog, and then, their posts are aggregated via RSS to the master weblog. Also, I don't know if there is people using this tool on a public site. I'm definitely interested to see a working example.

Update #2 [February 13 2003]: I think Dan Gillmor, although he is not used it about blogrolls, have a perfect expression to describe my feeling: echo-chamber effect:

One of the big problems in the weblog community is the echo-chamber effect, where like-minded bloggers link to each other and exclude opposing viewpoints. This is an especially serious flaw when it comes to politics, though I should note that some of the best political bloggers do invite counter-arguments.

comments: 0

10:25AM EST [ /Review | # ]