Metablog

Review of Crowston and Williams Paper

Summary:

This paper documents the range of communication genres found in a random sample of world wide web pages and attempts to identify new genres in the sample. According to the authors, the web is particularly interesting for genre studies because its rapid development encourages experimentation with new genres. There’s also the practical reason that many web pages are public, allowing a surveyor access to the pages for categorization. Last but not least, the web fosters communication between its many different users, and those users essentially bring their own familiar set of genres to the table, enabling a study of different genres and how they come together on the web to possibly form new ones.

Not surprising from the title of this paper, the authors explain genre change by breaking genres into two categories: reproduced and adapted. Reproduced genres are those that users already have experience with and now apply in a new setting. Adapted genres are those that have been modified to take on a new form and eventually become socially accepted as its own. According to the authors, the exact point of emergence of a new genre from an older one is hard to pinpoint. However, the important thing is for a genre to exist, it must be socially accepted in its communication setting.

In order to identify the genres in the WWW, the authors used Alta Vista and produced 1000 URLs of randomly selected single web pages from 1996, from which 837 actually existed by the time they studied them. They came up with a list of genres and subgenres, the results are included in the Appendix. When categorizing the genres (especially the emergent ones), they focused much on the use of links on a web page. For example, homepages and “hotlists” have their own patterns of inward versus outward links.

Finally, the authors state a practical use of their research results for web designers. In their opinions, web designers should draw from accepted web genres in order to better a reader’s ability to identify a site’s communicative purpose. They also make some practical suggestions on web page design for a single page.

Comments/Criticism:

In general I found this article to be well written and its discussion of web genres relevant (even if some of the results seem outdated). I just have a few comments.

I thought it was strange how the authors don’t acknowledge what I see to be a relatively big problem with attempting to identify web genres: the web is constantly in flux. It’s like having a moving target. By the time they went through their sample and categorized everything, the sample was already outdated as new genres emerge and old ones die. Blogs may not have existed when they took their sample, but it’s certainly its own genre now. It’s a good exercise, but I’m not sure the results are all that fruitful for the effort it took. Perhaps this kind of thing is perfectly accepted in the research community though.

Related to above, Yates and Orlikowski's paper (which this paper references quite a bit), suggests that "diachronic analysis is essential to observng the processes of genre emergence". Seems like these authors only focused on a snapshot in time, and therefore, may be missing the insight a study over a longer period may bring. Perhaps a more interesting study may be to take various samples over time and see how genres emerge or disappear (and the reasons for such changes).

Also, any attempt to list genre hierarchies will invite dissent. For example, should film reviews be listed on the same level as product reviews? or should they be categorized under product reviews? Some of their choices are more arbitrary and based on personal opinion (they acknowledge dissent even between themselves).

And lastly, do genres have to be adapted from existing genres? Or can completely new genres be produced in a communication setting without any precedent? I’m not sure if the authors really answer this question themselves, but I’m curious what you all think.

Posted by sttang at January 20, 2004 06:02 AM