Internet Archaeol. 15. Holtorf. Long-term preservation

3.3. Long-term preservation

My final worry was (wisely?) not discussed in the thesis itself, but was nevertheless very much on my mind at the time of writing (Holtorf 1999). That worry is how to archive electronic works permanently and reliably so that future generations of scholars will still have access to them. This issue of longevity is at the heart of traditional scholarship and it may also have to be rethought after the arrival of electronic communication (Burk et al. 2000; Stille 2002, chapter 11; Vielitz 2003; see also Richards this volume). In the case of my dissertation, I submitted a CD-ROM to my University. But although the relevant bodies had agreed to that, I knew that there would be problems ahead. It would be foolish to believe that current hardware and software standards are not going to change in the future. Even if HTML may still be readable in a decade, I doubt that the CD-ROM format we chose will be readable for a very long time. It is hardly surprising that the age of what might be termed 'electronic rescue archaeology' has already begun. PAD, an American preservation project, for example,

'seeks to identify threatened and endangered electronic literature and to maintain accessibility, encourage stability, and ensure availability of electronic works for readers, institutions, and scholars' (PAD Mission Statement).

The day will come sooner than many may imagine that some student or colleague will be curious enough to want to read my original dissertation and find out that this has actually become physically impossible. On one level, this is more a problem for the Library of my old University than it is one for myself. As far as I am concerned, the dissertation has fulfilled its main (though not exclusive) purpose of getting me a Doctoral degree and satisfying those who gave me financial and other support. Regarding the content of my work, it is about memory and multiple interpretations of ancient monuments anyway! What will remain are some spolia that will make their way into other works, either by chance or by deliberate selection. Curiously, works of electronic scholarship from the 21st century might thus turn out to survive in similar forms as the works of ancient scholarship from the Classical period (Stille 2002, 308-9). If traditionally 'a scholar is just a library's way of making another library' (Dennett 1992, 202), an electronic scholar is able to field some resistance to the expanding libraries.

Even if the updated and extended version of my work on the World Wide Web will be accessible for longer than the CD-ROM, this is not really of much use to those who want to check a quote in my original thesis. As I already indicated, my work is a 'living text' and things change. My Canadian publishers, CITD-Press in Toronto, now also sell CD-ROMs with copies of the current e-version at the time of production, but although each CD will have a specific date they will not all be archived by all libraries. Hence there will be some copies of older versions of my work somewhere, but it may be difficult to locate them. This is markedly different from Ted Nelson's vision of an electronic publication system in which 'a thing once published [will] stay published' (Nelson 1993, 2/43). Nelson argues that 'this is vital because of the links other users may have made to it'; the only way to incorporate corrections and amendments is by publishing superseding documents. However, since any reader 'will ordinarily want to see the new version' even links to an old version will normally be redirected to the same passage in the most recent version, 'if it's still there' as Nelson hastens to add (ibid.).

The difference then between Nelson's original vision and my own realisation is that in his case, readers can trace their way back to old versions of the same document while in my case they cannot. My own project therefore works more like the World Wide Web than like Xanadu: you always reach the most recent version, although sometimes this may mean that your link leads nowhere or not where it was once meant to lead. My own justification for this is mainly pragmatic:

It is far more important to read my current view on an issue than to find proof that I once thought otherwise. I may, for example, accept a criticism by a reader, change my text accordingly, and the reader's reference will henceworth not make much sense — but this is still better than leaving outdated views unchanged;
It is not worth keeping countless numbers of outdated pages in which only minor details, such as spelling mistakes, have been changed or to which a reference has been added — although this is by far the most common way in which I update my documents;
Were I to keep copies of every old version of all documents, my e-monograph would not only rapidly expand in size but also become very unwieldy and aesthetically clumsy — losing some of its beauty and elegance;
Not doing this is also the most efficient way of operating, since the vast majority of readers will want to understand my own position and argument and do not worry whether at some point I changed a comma or added a reference;
Those who do check up references by others to my work should extend the same trust they are prepared to invest in the presentation of my research results to the authors of the other texts they have read: if somebody quotes my work, but this quote has since changed or disappeared from it, it may be assumed that I once said so but would no longer like to say so now.
Should the situation arise in which I considered it worthwhile to keep an outdated version of any particular document available online, I am still free to do so — but this has not happened yet.

There is another criticism that can be made against my politics of continuous updating. Seamus Ross (2000) argued that we owe future historians very careful consideration of the preservation of fragile digital information. While his gloomy warnings about the possible loss of parts of contemporary scholarship and many digitally stored databases to future historians may well be justified, I do not agree with his logic according to which assumed preferences of future generations for particular kinds of source material should inform our actions now. Surely, it is not the task of future historians to judge whether we acted rightly or wrongly today but to understand how, and explain why, we did what we did. And presumably it will not be the quantity of available data that will determine the quality and significance of future historians' work anyway. No doubt future historians will take care of themselves.

Mark Aldenderfer (1999) found it somewhat disconcerting that I am not too concerned about the availability of my original Doctoral dissertation for future generations of scholars. But its potential future loss takes nothing away from what I have achieved when I wrote it, or from how I may or may not have influenced my contemporaries in their own respective projects. For some, academic archaeology is a long-term effort in which archaeologists, on behalf of all humankind, strive to complete knowledge about the human past; one building stone rests on another and before each contribution counts it has to be found acceptable by some form of peer review (Burk et al. 2000). Others see academic debate as a particular discourse about the past and its remains in the present, where entire buildings are being scrapped and rebuilt by new generations of architects and engineers. My e-monograph is written in a format that will most likely prevent it from being valued very positively by those agreeing with the former view. It is primarily a contribution to the current discourse of archaeology. The fact that my own work and similar transitory projects elsewhere already exist, with more in progress as we speak (Kirschenbaum 1996; and see http://directory.eliterature.org/), means that any future decisions have already been pre-empted: facts have been created now which everybody has to live and work with, or not work with, in the future.

At the Glasgow conference, much was made of the problems with long-term preservation of digitally recorded and/or published primary data (e.g. in Julian Richards' and Diana Murray's contributions and the subsequent discussions). Although all present academic discourse relies to a large extent on previously published data, and my own work is no exception, it is also true that primary data in databases does not exist somehow 'outside time'. Instead it is influenced by all kinds of value judgments and cultural constraints (see also Nick Ryan's paper). What once counted as relevant data to collect, how that data was collected, edited, presented, and accessed, and what kind of questions a researcher may ask at different times, all depends very much on changing conventions of academic discourse as well as on the approach and viewpoint of the particular archaeologist concerned. In this sense, even the collection of primary data is not necessarily cumulative in a straightforward way.

I have visited several times the enormous second-hand bookstores in Hay-on-Wye, with long shelves full of obscure archaeological literature of past decades. It always makes me feel guilty for even contemplating publishing yet more works. To me, many of these under-appreciated volumes, available for a few pounds, are monuments of the failure to realise in time that ultimately maybe they should never have been published and that countless of the authors' evenings and weekends might have been better spent doing things other than writing them. I am therefore pleased that although my own monograph may gradually become redundant too, at least it will never stand on a shelf longing for attention, desperately seeking lovers. The files that make up my work will never become a 'standing monument' for their own redundancy or for the underlying misjudgements by which they may have been written and assembled. Instead, my monograph will disappear into the endless ether of information, i.e. precisely where it came from. And many of the circulating CD hard copies may find their final purpose in enriching the stratigraphy of local landfill sites. A Monumental Past? Hardly.