Websites as Stable Resources

Harrison Eiteljorg, II

(See email contacts page for the author's email address.)

Reviews of websites of interest to archaeologists have been carried in the CSA Newsletter since February of 1996, and the Bryn Mawr Electronic Resources Review was established to review similar websites, with the first such review appearing in June of 1998. Although BMERR no longer exists as a stand-alone review site, reviews continue to be a regular feature of the CSA Newsletter. Reviewed websites available via CSA and BMERR total 60 as of October, 2011, with some reviews covering more than one website and two sites reviewed twice (as a result of which the numbers cited here may not seem right at times).

Professor Richard Hamilton, a member of the CSA board and co-founder of the Bryn Mawr Classical Review, recently suggested a look back at the websites that have been reviewed in the two publications to see how many remain available and what changes have occurred. That seemed a good idea, and what follows is a review of those websites, not in the sense of the original reviews but to obtain a general sense of whether websites have or have not been maintained and other issues that may seem to be notable in the examination. This is not a second (or third) look at any of the websites; there will be no examination of the content here.

The first question is the simple one. How many of the websites are still available, either at the URL used at the time of the review or via some other route? As one might expect, the more recently reviewed sites are the most likely to remain available. In fact, only two websites of the 24 reviewed since the end of 2004 are now unavailable. One of them, the Archaeological Atlas of the Aegean — at samos.ypai.gr/atlas/Default_uk.asp, may be unavailable because of economic issues in Greece. (Access was first unsuccessfully tried in August, 2011, and last unsuccessfully tried in January, 2012.) The other, a website for an excavation, has left little trace. A third website, the one for Çatalhöyük is now on its second URL. While the website is easily found, it does not seem to be too much to ask that an old URL forward users to the new one. One other site required the addition of "www." to the front of the URL, but there was an automatic redirect on the page without the addition; using the old URL still worked.

In one case — the Archaeological Institute of America's Archaeological Fieldwork Opportunities Bulletin listing on the web — the page no longer exists at the original URL but can be found with relative ease via the AIA's website. It suggests the kind of inattention too often found when major changes to a website are undertaken. A redirect page should not have been omitted for such an important page, especially one that had been reviewed.

Only a few reviews, for some reason, were published from 2000 to 2004 — three in 2004, one in 2002, none in 2001 or 2003. Of those four, one could not be found without some effort. The other three URLs were all still effective.

As we look back at sites reviewed before 2001 (33 reviewed sites), the results are quite different and very unsettling. Only twelve of those websites loaded without a problem (including some that automatically redirected users to a new site, one that provided a link to the new site, and one that required a slight change in the URL by me). Of the others, some were relatively easy to find via searches, but nine could not be found at all, and more than one of those found did not seem to be the website that had been reviewed. (Note that one of the sites not found via the URL in the review was the website for Çatalhöyük. The URL had changed, but there was no link.)

As I looked at these results, I wondered whether a careful user would be able to rely upon what he/she found at these sites, assuming a given site could be found. That is, when continuity is unclear (and it may be unclear for any or all websites with inconsistent URLs if the history of these is indicative), what other clues are needed to let the reader know whether an individual web page is to be trusted? For me the answer is simple, and the CSA website has been constructed with my views in mind. Every web page, with rather few exceptions,¹ has the author clearly stated. In addition the date of publication is noted either in the About section or, in the case of newsletter articles, in the header. Other information, including a history of revisions, often quite brief, is included, and some documents have prior versions retained, with URLs provided. That level of documentation seems to me to be required, but it is not regularly available on the websites that have been reviewed. I have not examined all websites reviewed, but I have looked at those that have had a difficult history with some care, the ones that were hard to find. It seemed to me that those sites truly needed authors and dates of authorship to provide some comfort to users who could not be sure they had even found the old website they were seeking. In fact, though, I found one and only one site that included the author's name and the date on every page (that is, every page I examined). Indeed, many websites were not certainly being maintained since there was so rarely any information on the home page to make it clear that the page had been updated recently.

In general, it therefore seems that websites have often been constructed with little or no regard for their long-term viability or utility. (I note here that some bibliographies available at these older, inactive websites give some sense of date if one checks the latest included publication, but that is a very poor way to determine a date of writing.)

This is not an encouraging finding. It is even more discouraging to realize that many of the sites that do not function now had URLs showing that the sites had been hosted at universities when first established.

There is, unfortunately, an important lesson a user may take away from this information. In short, be very wary about trusting websites. The ones you use may or may not be here tomorrow. And if they are still here, they may or may not be where they can be found or present information that still seems trustworthy. There is also a critical lesson for those running websites. If they want their sites to be valued and used, they need to be more explicit about the care of the materials on the site, their authorship, the dates of posting, the past and future of the site, and so on. Those websites that show clearly and explicitly that their creators are aware of these issues seem to me to be saying that their work can be trusted. At this point, though, it is hard to imagine trusting an academic website that does nothing to show to its readers that the stability of the site itself and the reliability of its content are issues taken seriously by those operating the site.

-- Harrison Eiteljorg, II

Notes:

1. CSA Newsletter articles were not attributed to me in the early days, when I wrote all of them. As time progressed, more articles were attributed, but I often did not attribute my own articles to myself, and that was especially true when Susan C. Jones was working with me on those articles. It seemed hard to assign authorship in some cases. The home page for the CSA Newsletter now states that any article without a stated author should be assumed to have been written by me. Return to text.