If perchance I come up with something more insightful to comment, I'll trackback it, but for now my comments and questions are as follows:
1) I really enjoyed the paper generally and the four (and a half) major points. The forensics angle is productive, though I can't stop thinking about William Peterson examining my computer.
2) Do you ever address the process of data corruption on disks, or (as a related topic) the influence of file systems in the representation of data? This "corruptibility" certainly could have something to do with the ephemeral feeling of computer data, though it actually is not terribly different from a piece of paper getting crumpled or wet.
3) At some pint someone (like me) really should finish CodedAndRecoded because if it ever got done it would relate nicely to the various levels of permanence of the data.
4) Palm just recently released a PalmPilot where the Flash memory is non-volatile; it can lose power and hold its data on the chips. I wonder if at some point RAM may go this way too -- so that power-outages wouldn't matter (as much) to a computer's operations. Either way, it certainly cuts into idea that data on chips is less permanent.

Thanks, Matt. Exactly the kind of feedback I'm hoping to get.

And maybe you (or I) could talk up CodedAndRecoded to the local digerati.

Just a quick note for an interesting paper:

Do you see any potential confluence between your handling of digital materiality and traditional manuscript study? I know you've elsewhere advocated the importance of traditional textual criticism for digital media; here I think it could be a productive relationship simply because many manuscripts exhibit erasures that aren't reallly "gone," much like "erased" data on magnetic media.

My feet are barely wet in terms of humanities computing and digital scholarship, but I'm excited by the possibilities these things pose, especially in my work with Old and Middle English texts. Thanks for your inspiring work.

Erik Vorhes
Loyola University Chicago

Thanks for the kind words, Erik. This paper was partly written under the sign of the Folger seminar I recently attended on early modern writing technologies, so I see lots of commonality with codicology and traditional concerns of textual studies. The search for an erasable writing technology, which has been a constant throughout the history of writing, carried over directly into computer storage media, for example. At the same time, I'm always cautious about pushing the similarities too far--you run the risk of failing to respect the distinctive materiality of either medium that way. Mangetic tape is *not* parchment or rag paper, and each exhibits their distinctive properties.

For me the most important issue is not necessarily that the bibliographer learn his or her way around a scanning electron miscrope, but that by debunking some of the myths surrounding electronic writing people may be more proactive about understanding the importance of doing what they can locally to assist in its preservation. Keeping an old hard drive, for example, even if you're throwing away the rest of the computer. Because the bits *are* still there, and will be for quite some time.

Interesting convergence on both of your points is occurring as we speak at Oxford University, where previously "unreadable" (apropos for both traditional and digital textuality) scrolls have been restored due to digital imaging. Another illustration of how difficult it can be to fully erase inscribed media (although a nice fire would have done the trick here, too). Here's the link to the article:

http://www.gnn.tv/headlines/2532/The_Beast_Gets_A_New_Number_In_Time_For_Karmageddon

Matt, I like how the paper begins almost in media res with an anecdotal approach. Rhetorically it provides a journalistic fact checking flavour to what follows. At first I was put off by what I took to be a derivative polemic [the wave to McGann's contention that certain literary theory is indebted to Romanticism}. I wanted to see a note referencing the books of the authors listed in the [naively anti-materialist?] camp. I just thoguht that the enthymeme that equated the "erasable" with the "ineffable" is understated and this reader's hankering for evidence distracts from your main point about transmission and inscription. However, your final word made it all come together -- a paean to the forensic! For example Bolter's remediation concept is indebted to McLuhan and is grounded in a belief in the impossibility of translation between sensory modes. A belief put to shame by even a cursory analysis of the encoding and decipherment at work in moving bits. On that count, your paper provides a nice position
to re-read David Weinberger's account of moving bits in the "matter" chapter of Small Pieces Loosely Joined. His is a dualist position but in one passage he has a felicitous expression that put me in mind of your praise for textual forensics [aside: why not trumpet the long tradition of document sleuthing coming down to us in the West from the Humanist challenge to the authenticity of the Donation of Constatine?]. Weinberger writes: "A bit is only a bit because humans have designed a system to take a bit as standing for something." The mark registered on/in a given medium is of course the inscription but there is also the inscription as an activity (a gesture performed in time). Those screen-conscious (all American?) theorists listed in your paper seem to fetishize this second notion of inscription (a gesture in a given time and place). Peirce's semiotics with the importance it places on communities of interpretation might provide a bridge? In any event, other communities figure Romanticism differently and
look to among others Shelley's poet as legislator or Holderlin's as read by Heidegger "Language is charged with the task of making beings manifest and preserving them as such [...]". Weinberger tries to keep "bits" and "words" in separate categories. I think your work on storage and retrieval shows that from a materialist semiotic perspective the distinction cannot hold. In short, your work in this paper shows (1) No transmission without inscription; (2) the computer as machine is actually an assemblage of machines connectable to other machines and that such machine connections make available the un(re)read yet still readable and that machines also destroy the un(re)read -- i.e. the electronic is not manifestly of a different metaphysical sort than the mechanical in the making and unmaking of readability; (3) the ineffable is not the erasable

Good luck with the book!

Thanks Francois, these are all great comments. Genuinely helpful.

I realized that I never gave a ref on that non-volatle palm, and so here one is:

http://www.palmone.com/us/products/handhelds/tungsten-t5/

Apparently they've put flash memory like this in all their new stuff. You may also be interested in their new product, the palm "Life drive," which is hdd based and is advertised interestingly:

http://www.palmone.com/us/products/mobilemanagers/lifedrive/

I don't think I'll ever find a real link on those shakey hdds, but I'll keep looking.