21.501 no killer app but work to be done

From: Humanist Discussion Group (by way of Willard McCarty willard.mccarty_at_kcl.ac.uk>
Date: Fri, 25 Jan 2008 09:48:50 +0000

               Humanist Discussion Group, Vol. 21, No. 501.
       Centre for Computing in the Humanities, King's College London
                     Submit to: humanist_at_princeton.edu

   [1] From: amsler_at_cs.utexas.edu (71)
         Subject: Re: 21.496 no killer app but work to be done

   [2] From: Michael Hart <hart_at_pglaf.org> (23)
         Subject: Re: 21.495 no killer app but work to be done

         Date: Fri, 25 Jan 2008 09:23:37 +0000
         From: amsler_at_cs.utexas.edu
         Subject: Re: 21.496 no killer app but work to be done

The difficulty with finding a "killer app" is that you're asking for objective
results rather than subjective results. I do not anticipate a mechanism for
objective results in literary analysis, nothing like the discovery of carbon-14
dating for physical paper.

What things do seem to divert the course of events in Humanities Computing?

First and foremost are new FREE resources. Project Gutenberg certainly changed
the course of humanities research, providing etexts of classics. It also
provided a mechanism for large cultural organizations (like Canada!) to promote
their historial literature worldwide with minimal cost. Whenever resources are
made available for free they promote grass-roots investigations that otherwise
wouldn't take place. The Linguistic Data Consortium preserves data, but hasn't
had the same sort of impact because its data is so expensive, especially to
individuals without sufficient institutional backing. The Oxford Text Archive
is somewhere in the middle, with modest cost requirements. I think there is a
direct inverse correlation between the cost of resources and the impact of
digital information. Google is a strange blend of free and proprietary data
such that its well-publicized projects (Google Maps, Google Books, Google's
N-grams) have some dramatic impact, but NO serious impact when you reach for
the data they don't release (e.g., the OCR text of Google Books in the public
domain released as PDF image files).

The FREE avialability of the Wiki software also appears to have had a
significant impact on what can be done. This impact is muted only by the
relative difficulty for non-systems programmers in using this FREE software.
What would greatly alter that would be for someone to offer a free downloadable
plug-and-play Wikimedia software system for Windows PCs. Download this, run
this, and you can create a Wiki on your system. I personally believe Wikis are
the most significant non-Microsoft Office killer app of the last several years.
Wikis realize in practice what Apple hinted at in hypercards, but failed to
achieve because they couldn't scale upward to users expectations.

Lacking FREE resources, the next best thing are additions to nearly universally
used commercial software. The first "killer app" was spreadsheets in
Microsoft's Excel. The reason that "app" was a killer probably had more to do
with its inclusion in the widely distributed standard Microsoft Office suite of
software than with its perceived utility. If spreadsheets had been sold as a
stand-alone product from XYZ-corporation, marketed to accountants and costing
$1000 each unit; it wouldn't have been a "killer app". Anything that gets added
into a nearly universally available software utility can have significant

Customized software packages that require specific purchases for each
are limited in what they can achieve. This is a software analogy with
a standard
axiom of hardware development. Special purpose computers cannot compete
long-term with general purpose computers. The cost of special-purpose products
is not contained in the one-time creation of the product, but in the lifelong
maintenance and upgrading of the product. Usually, the general purpose
product's upgrades will completely overtake the special purpose product in a
couple of generations---and generations are only 18 months apart.

In software the product never exceeds its targeted market, and hence can't
achieve "killer app" status because that requires the general computing public
to find applications for the software not anticipated by its developers.
The fact that a spreadsheet can hold historical data, literary references,
compile footnoted references, build indexes, etc. are all beyond its envisioned
role as an accounting system utility.

So, there you have it. I think the Humanities should pour labor into projects
such as Project Gutenberg. While the publishing lobby managed to block public
domain texts from extending after 1923 for the rest of our lifetimes, that
still leaves not only a vast amount of text before 1923 to be captured but the
prospect of obtaining permission from post-1923 authors for submission of their
still-copyrighted works.

Project Gutenberg however needs help becoming more authoratitive (just as Wikis
need help becoming more scholarly and reliable). The Project Gutenberg aversion
to identifying specific editions of texts is damaging to the ability to use the
data for more comprehensive digital computing. It is highly commendable that
Project Gutenberg is exemplary in their standards for accurate text, but
lacking the ability to know which editions of a work are being represented
limits the applications of the data. Once again, it is not the perceived goal
of the developers (making texts available to human readers) but the ability of
the data to find additional uses (e.g., corpus linguistic analyses) that
distinguishes good resources from broadening their appeal to the level of a
"killer app".

         Date: Fri, 25 Jan 2008 09:28:06 +0000
         From: Michael Hart <hart_at_pglaf.org>
         Subject: Re: 21.495 no killer app but work to be done

True, you can't convince the skeptics. . .you still can't say
that digitial music has wiped out analog music because a few
places still make analog records which are really better, not
that a true skeptic needs those last few words.

Even when there are more eBooks than paper books, no way.

Even when there are 100 times as many eBooks, not happening.

It's not going to matter what they SAY about eBooks, reality
is going in that direction and paper books will never reverse
that trend, simply because you can /OWN/ MILLIONS OF eBOOKS IN
A TERABYTE DRIVE [costing under $200].

Before Gutenberg the average person could own zero books.

Before Project Gutenberg an average person could own 0 libraries.

It's literally as simple as that.

The cost/benefit ratio for eBooks is too much better than paper.


Michael S. Hart
Project Gutenberg

Recommended Books:

Dandelion Wine, by Ray Bradbury: For The Right Brain
Atlas Shrugged, by Ayn Ran,: For The Left Brain [or both]
Diamond Age, by Neal Stephenson: To Understand The Internet
The Phantom Toobooth, by Norton Juster: Lesson of Life. . .
Received on Fri Jan 25 2008 - 05:18:12 EST

This archive was generated by hypermail 2.2.0 : Fri Jan 25 2008 - 05:18:12 EST