?

Log in

No account? Create an account

Previous Entry | Next Entry

Steganography and Paper

I wrote a paper for the First International Conference on the Future of the Book in 2003. It was published in some of the conference apocrypha, and I used to have a link to a copy online, but it appears to be one. If I can dig up the proof later, I will repost it.

In that paper, I vigorously defended the need to ensure total portability and maintainability of a digital library. Countless of times over the course of history thought has been pooled by the destruction of a single tome which contained unique thought of the time. When the physical book was gone, the thought was erased, or obscured until rediscovered. Recent work on some of Archimedes' texts have shown that the ancient Greeks (or at least one of them) knew a whole lot more than we give them credit for. Digital libraries need to have universal form and function, so they can be integrated into a greater collective of knowledge.

My paper argued that the meat of the book, the words, could be captured electronically and flawlessly. This process would preempt the happenstantial destruction of a work, and allow for simple and fast replication and dissemination. Further, with all works eventually becoming digital, cross referencing, indexing, and knowledge slicing would become streamlined to the point where you literally could find the connections between anything.

In Matthew Woodring Stover's Heroes Die, we are given a fantastic view of what things may become (albeit with an incredible fantastic twist). The idea of purely electronic documents allows for the control of the text, and the ideas behind it. Books are tactile things, and hard to edit seamlessly. Electronic documents are susceptible to revisions without it ever being obvious to the reader, and in certain scenarios, even real time. The use of editions in large-scale publications address this somewhat in terms of being able to track changes across the history of a body of work in paper, but no such device exists for electronic documents in eprint.

The father of the protagonist in the book points out that if you have a printed copy of a book, nobody can twist or change the knowledge in that book without you knowing it. It is a particularly poignant point in the face of the big-brother information control environment, but one which I ultimately threw by the wayside as alarmist claptrap. After all, couldn't people have been doing such things for years with conventional printing? From what I know about the bookmaking industry, enough things get into works gone to press that are _not_ supposed to be there that it wouldn't surprise me if someone wanted to play big brother with a white pen, they could do so with relative ease. Perhaps naively, I assume that there will be no one single controlling interest overseeing the entire digital library process. That was why I was so adamant about open standards, so everyone is doing everything the same way, despite the fact that it was not all being done by the same people.

Many of the detracting arguments against eprint are aesthetic. Typeset, font, paper's tactile value and color. The power of good binding and strong cover (not to mention cover art). All these things, to me, are secondary to the actual work - the words which capture the writer's intent and pass them on to the reader. Some people can't get past the actual process of flipping pages and holding a sheaf of paper. To them, I usually say, get over it.

Today was the first time, however, that something occurred to me which was a spot-on argument against eprint.

Steganography.

Not the new fangled hard core digital pixel manipulation encoding we all know and love thanks to media mongers after 9/11. I'm talking about the old school stuff. The stuff of the 14th and 15th century, which _I_ know and love.

All old textual stegnographic works that I have ever done any work on are all contextually based. That means that font type and placement, graphic placement and where that graphic related to the page count, chapter breaks etc, as well as the spacing of columns and tabs were absolutely integral to the ability to decode ANYTHING. Granted, some works (like Trithemius' stuff, and Wolfgang Heidel's writings) can be brute forced mathematically if you have the sequences and apply the correct pattern.

But, for true textual steganography, there is too much contextual nuance to be captured digitally, unless you do image scans. Image scans are something I am totally against, because of quality control issues, storage size, and the inability to tag them in a reliable manner with minimal human interaction (and therefore chance of error).

What this has made me realize is that most books are soulless things. They are little more than sentences and paragraphs of sinew and meat slapped over a skeletons of chapters and sections. However, in certain cases, the actual artifice of the book gives a physical work a soul. This is particularly the case in pre-printing press works, but there are clear exceptions to that generalization.

When the soul of such a book is to encapsulate a hidden meaning or message, the translation to electronic media in an elegant format seems too daunting a task to contemplate. Combined with the actual loss of the artifact itself as the de-facto source of experience (see the Archimides link above), I am now beginning to wonder if perhaps I need to revise my outlook on the significance of implementing true digital collections.

All this on less than three hours sleep! What have you done with your morning?

Comments

( 4 comments — Leave a comment )
gemini_heart
Jun. 30th, 2005 04:18 pm (UTC)
Allow me to play devil’s advocate here…
I view e-publishing as a desirable and viable alternative to traditional publishing methods, not a replacement. It is nice to theorize about a digital library, but it would then fall to some organization to maintain and regulate that library, unless you want to leave that task in the realm of private industry (i.e., under the purview of the publishing companies themselves). However, if you want to have standards, you still need a governing body to decide upon and enforce those standards.

My paper argued that the meat of the book, the words, could be captured electronically and flawlessly. This process would preempt the happenstantial destruction of a work, and allow for simple and fast replication and dissemination. Further, with all works eventually becoming digital, cross referencing, indexing, and knowledge slicing would become streamlined to the point where you literally could find the connections between anything.

It seems to me that it would be easier to destroy a digital library than it would be to seek out and destroy all the thousands of hard copies that exist. Also, the effort involved in digitally cross-referencing and indexing millions of books is mind-boggling. The creation of relevant cross-references and indexes is not a task that can be left to computers; it takes human intelligence to interpret the context and determine relevant entries.

Electronic documents are susceptible to revisions without it ever being obvious to the reader, and in certain scenarios, even real time. The use of editions in large-scale publications address this somewhat in terms of being able to track changes across the history of a body of work in paper, but no such device exists for electronic documents in eprint.

Change control and comparison software could be used to ensure no unauthorized changes are made to publications. It should be relatively easy to lock down words the same way you can lock down code.



delascabezas
Jun. 30th, 2005 08:14 pm (UTC)
hehe, devil away
as to destruction: yes and no. if you think about it, the tragedy of the burning of the library of alexandria was tragic because of the number of irreplaceable works destroyed. now, if you could fit the contents of the library on to say, something the size of an ipod, it may indeed be easier to destroy, but it is MUCH more portable, and very easy to replicate and back up. the odds of ther eonly ever being ne ipod anywhere at any time with everything on it are way lower than "last print editions" and "loan library collections" getting robbed or destroyed.

as to the cros sindexing, i would propose that for now youuse a semi-intelligent algorithm, until artificial intelligence is truly a reality. once that comes into play, you have something out of star trek, where you can query the same system for any kind of data.

i agree wholeheartedly we are not at the place and time where traditional publication can be unseated by eprint, but that is defintiely the direction things are going in. one of the best papers in the converence talked about vellum preparers in guilds in venice dismissing offhand rag paper, only to find themselves unemployed within a decade as a result of the printing press. humanity is good at droppping one thing and runing to another, provided there is enough incentive. i think the possible benefits of going paperless are there, the technology just needs to catch up.

the standards i propose would be governed by the body using them - that is what open standards are all about. sure, there is always some consortium administrating them somewhere, but they don't make the decisions per se; they just formalize what the majority concencus is, and poke the keystone log in the case of a log jam.

the last item is the biggest bone of contention i have with ANY centralized electronic system. human nature has proven again and again the corruptive nature of power. if someone were able to hack etexts, for fun or profit, then the entire thing becomes useless. without the actual printed text to fall back on, there would be nothing to stop someone from changing works, particularly after a generation or two got used to not questioning what they are reading. in this canse, the fear that it might happen is almost enough to prevent it from happening, if eprint is going to become the medium over traditional paper.
sophiaserpentia
Jun. 30th, 2005 05:01 pm (UTC)
Hmm. The creation of new forms of data storage might make the file-size issues with regard to digital imaging moot in a few years. But that is the least of your points against it.

The problem though that you refer to is just a projection of the difficulty you get with writing things down in the first place: it is like trying to capture the fullness of a flowing river in a still photo. Digitalizing that which is intended to convey information contextually loses that information, just as writing something down causes you to lose much of the resonant cultural context of speech (that loss becomes especially acute over time). I don't know how to counter that, other than for some of us to be steganographic "luddites."
delascabezas
Jun. 30th, 2005 08:06 pm (UTC)
mm
this is more of a central issue of epistimology though - the meaning of the author vs. what the reader brings to the work. as far as stegnography is concerned, the book is the medium through which the author is creating the flowing river; quite a puzzle.

i agree that written words have limitations, bt we live in a world with too much information to return to an oral tradition. maybe once technology or biomechanics gets to the point where we have limited telepathy, it will be possible. until then, i think we are stuck with what we got though.
( 4 comments — Leave a comment )

Profile

Steam Escaping!
delascabezas
The Son of the last of a long line of thinkers.
delascabezas.com

Latest Month

February 2017
S M T W T F S
   1234
567891011
12131415161718
19202122232425
262728    

Tags

Powered by LiveJournal.com
Designed by Tiffany Chow