I appreciate being given access to this list, which is furthering a
discussion of importance by important players. As I was only notified of
access on Friday, my comments on the advance papers for the conference
(starting today, Sunday) may be of minimal impact, for as I write you are
sitting down to dinner at the Getty. I will therefore be brief in this first
message (about 2 typed pages).
Three comments, each of which will be expanded upon:
1. The risks of the holistic approach;
2. The risks of overdependence on market rhetoric;
3. The need for attention to integrity, or authenticity.
1. The holistic approach: the documents, particular the early part of the
background paper by Lyman/Besser, refer too apocalyptically to the loss of
digital information. Similar statements could be made about the loss of
written and printed documents (to say nothing of oral culture, whether Greek
or present-day). Vast quantities of information are destroyed daily and
always have been, whether in wastebaskets, shredders, or to start cooking
fires. Overdone statements about current loss that don't relate to our
experience with paper lose credibility and defeat analysis.
There always have been mechanisms for the saving of what is important, and it
is those that we need to replicate digitally. The mechanisms are not
perfect; we are fortunate that Pepys decided to save ballads (now at
Cambridge's library) but we don't know what genre or body of work of other
kinds we would have if someone had saved them--for they didn't. But there is
probably a very rough correlation between what is considered to be important
and what has continued to be saved.
This is primarily because the saving process is multivariate and
micromanaged. Individuals or small organizations have saved what is
considered to be valuable according to their own judgements and procedures.
It is true that technologically we need to provide relatively better tools
than existed in the past so that digital information can be saved. But it is
not helpful to confuse the technological need with the sociological
likelihood of very distributed use of the mechanisms. Or, to put it another
brief way, everything won't be saved and we need to accept that.
2. The market approach: the Lyman/Besser paper assumes the rhetoric of the
free market where it needn't (e.g. III.2, III.4). (IV.5 is a striking
example: "...the concept of public good is being replaced by the
commodification of digital knowledge...." The statement's passive voice
accepts the market mechanism as a fact of nature, when it certainly isn't.
In this case, particular enterprises are acting to replace public good by
commodification, as is usual in the marketplace. Other enterprises (mostly
but not all nonprofits) are acting to resist that replacement, and in fact
have done so in varying ways for some time.
The current market zeitgeist (and the background of many conference
participants) may encourage such thinking, but it hasn't driven the
preservation of the human record in the past and -- if we are to succeed --
it won't in the future (libraries are not economically sound investments nor
are they cost-effective in any narrow sense). Conscious, planned decisions
must be made based on values other than perceived current or future economic
return, just as they have in the past. The paper gets at this in paras. II.4
and II.5 in which there is a nod to cultural factors as well as economic.
In para. II.4 the question is asked, "will institutions originally
established to house print documents be robust enough to act as sites for
digital archives?" By "institutions" I read "libraries," for with archives
they are the only ones who do this. The question is well asked of libraries'
technological capability, for they are only just coming up to speed. But the
thrust of the question is organizational, and here the answer has to be that
we must assure that they are robust enough to do so, for if they do not we
either will fail in preserving the digital record or we will create
organizational structures that look remarkably like libraries and archives.
This is not, I hope, a narrow defensive reaction from a librarian, but a
statement of the case: the social institutions which are charged with the
preservation of human informational marks are called libraries. Our stance
needs to be that we *will* make them organizationally robust enough. (I
assure you I am not defensive enough to say that it won't be difficult.)
3. Intellectual preservation: the papers created for the conference make
almost no reference to one of the genuinely new problems in digital
preservation: the need for authentication (or integrity or intellectual
preservation). In the print environment, a printed text was difficult to
change or tamper with -- the text you cite is certain to be the text I read
even from a copy on a different shelf. In a way, the digital environment is
a throwback to the manuscript world: I have no assurance (yet) that my
digital copy is the same as the one you used yesterday (or last year or in 50
years). The consequences for scholarship and public debate, not to mention
creative endeavor, are enormous.
The Lyman/Besser paper does mention "authoritative versions" (IV.4), but not
in the context of assurance. It is essential that the concept of
preservation in the digital environment include the notion of integrity, or
the nature of what is preserved will be suspect from the start. There are
tools and methodologies available and some thought has begun to be applied in
this area. See (i.a.) Graham, Peter S., Intellectual Preservation:
Electronic Preservation of the Third Kind (Washington, DC: Commission on
Preservation and Access, March, 1994), 8 pp.
<URL:http://aultnis.rutgers.edu/texts/cpaintpres.html>; and Lynch, Clifford,
The Integrity of Digital Information: Mechanics and Definitional Issues,
Journal of the American Society for Information Science, 45(10):737-744,
1994, as well as other items in the conference's own web bibliography (I
offer my own bibliography on digital preservation at
<URL:http://aultnis.rutgers.edu/texts/ElectLibBib.html>.)
Thank you for listening. --pg
NEW ZIP: Peter Graham Rutgers University Libraries
169 College Ave., New Brunswick, NJ 08901-1163
<URL:http://aultnis.rutgers.edu/pghome.html>