7.0144 Wittgenstein in Electronic Facsimile (1/155)

Mon, 30 Aug 1993 10:47:00 EDT

Humanist Discussion Group, Vol. 7, No. 0144. Monday, 30 Aug 1993.

Date: Mon, 30 Aug 93 08:42:48 CET
From: "Claus.Huitfeldt" <Claus.Huitfeldt@EDB-hum.NAVF.pc.hd.uib.no>
Subject: Wittgenstein in Electronic Facsimile

Bergen, August 1993
- report from a feasibility study

The Austrian philosopher Ludwig Wittgenstein (1889 - 1951)
published only one book of philosophy, the Tractatus
Logico-Philosophicus. On his death, however, he left behind
approx. 20,000 pages of unpublished manuscripts. Many of those
(e.g. the Philosophical Investigations) have been published
posthumously, but even today large parts of his writings remain

The Wittgenstein Archives at the University of Bergen aims to
make all the writings of Wittgenstein available in the form of
machine-readable transcriptions (tagged text files). Approx.
seven thousand pages have now been transcribed. Finishing the
task is expected to take several years. During November and
December 1992 a feasibility study was carried out at The
Wittgenstein Archives to consider the production of a facsimile
CD-ROM version of Wittgenstein's collected writings, to be
published by Oxford University Press prior to or in addition to
the transcriptions. The following is the summary of the report
from the feasibility study:

The aim of this feasibility study was to provide technical
specifications for and assess the costs of the production of a
facsimile CD-ROM version of Wittgenstein's Nachlass, or, as we
have called it, WEF (Wittgenstein in Electronic Facsimile).

The study was carried out in the period 18 November to 17
December 1992. The Wittgenstein Archives at the University of
Bergen (WAB) was responsible for organizing and implementing the
study. Representatives of Oxford University Press, the Norwegian
Computing Centre for the Humanities and the Department of
Research Management at the University of Bergen also took part
in the work group.

Wittgenstein's Nachlass consists of 155 items of approx. 20,000
pages of manuscripts and typescripts. The typescripts (approx.
5,000 pages) and a few manuscripts are on loose sheets, the rest
in bound volumes and notebooks. Maximum page size is 30 x 46.5
cm, average page size is 80% of A4 size. There are problems with
low contrast, bleed-through and writing in or across the gutter.
Most of the original material is held by Wren Library at Trinity
College in Cambridge, a substantial part by the Austrian National
Library in Vienna, and the rest by the Bodleian Library in Oxford
and the Bertrand Russell Archive in Hamilton, Ontario.

WEF's target audience is university libraries, philosophy
departments, other research institutions, and individual
scholars. Its target technological platforms (in order of
priority) are (1) IBM compatible PCs with 386 CPU or higher,
Microsoft Windows 3.1 or higher, VGA or SVGA color screen and
CD-ROM player; (2) powerful Macintosh computers with QuickTime,
color screen, and CD-ROM player; and (3) UNIX-based systems.

Experiments with scanning of similar material and contact with
possible vendors and cooperation partners provided the basis for
the study. The main problems are on the one hand those of image
capture combining unharmful handling of the original manuscripts
with acceptable throughput and image quality, and on the other
hand those of keeping storage space requirements within a
reasonable number of CD-ROMs.

We have defined three possible target reproduction qualities for
consideration: Minimum, Medium, and Very High Quality. Minimum
Quality requires 72 dpi resolution, and 16 shades of grey. Medium
Quality requires between 72 and 250 dpi with an average
resolution of 100 dpi, and 256 colors. Very High Quality requires
between 250 and 400 dpi with an average resolution of 300 dpi,
and 16 million colors. Medium Quality is considered in most
respects superior to ordinary microfilm, and Very High Quality
in all respects superior to ordinary microfilm.

It is possible to create a Very High Quality electronic
facsimile, but it is not feasible for a publication project. WEF
should contain all images in two versions:

Minimum Quality, i.e. 72 dpi and 4 bit pixel depth (16 shades of
grey), compressed with GIF at an average ratio of 1:5, stored on
one CD-ROM disk.

Medium Quality, i.e. an average resolution of 100 dpi with a
minimum of 72 dpi and a maximum of 250 dpi, and 24 bit pixel
depth (16 mill. colors), compressed with JPEG at an average ratio
of 1:20, stored on four CD-ROM disks.

Most considerations, those of image quality as well as those of
total costs, speak in favor of direct digital scanning from the
original documents. However, direct scanning has one decisive
disadvantage: The effects of subjecting the original documents
to strong light for the amounts of time required by this process
are unknown and may be harmful. If further investigations should
indicate that this obstacle can be overcome, or one could be
assured that scanning is not more harmful to the originals than
photographing, direct scanning is highly recommended. However,
we have found no positive indications that this may in fact
happen in the foreseeable future.

We therefore recommend that the original documents are captured
on 35 mm color film, which is then digitized in a separate

An in house process making use of a slide scanner will provide
the necessary image quality, but is a relatively labor intensive
strategy. Experiments performed so far indicate that Kodak's
Photo CD may provide sufficient image quality.

These strategies both have the advantage of creating a 35 mm
color film version, and, with the latter strategy, also a
(180-disk) Photo CD version, of the entire Nachlass. These copies
are of higher quality than the distributed version and may
therefore be kept as archival copies for backup or later use.

As to the choice between 35 mm normal film and 35 mm microfilm,
the project faces a dilemma: Capture on microfilm is inexpensive,
but microfilm cannot be scanned in colors. Normal film can be
scanned in colors, but capture on normal film is expensive. We
believe that further investigations will provide a solution to
this problem, and that transferring 35 mm color microfilm to
Photo CD will be possible.

The project will also include the preparation of a data structure
and a user interface for accessing the images, based on von
Wright's catalogue to the Nachlass and WAB's registration
standard, MECS-WIT. It is important that WEF allows for inclusion
of transcriptions of the texts.

A 13 month project plan has been proposed according to which the
first 3 1/2 months will be spent developing a prototype. On the
basis of the experiences gathered during this process, the
methods and technology should be evaluated, possibly leading to
a reassessment and a revision of the budget. This plan includes
the development of a rather simple user interface only, and does
not include the suggested Macintosh and UNIX versions of the

| Claus Huitfeldt, |
| The Wittgenstein Archives at the University of Bergen, |
| Harald Haarfagresgt 31, |
| N-5007 Bergen, |
| Norway. |
| |
| Tel: +47-(0)5-21 29 50 (From Sept. 9, 1993: +47-55-21 29 50) |
| Fax: +47-(0)5-32 26 56 (From Sept. 9, 1993: +47-55-32 26 56) |
| |
| E-mail: claus@pc.hd.uib.no. |