ware such as MacPaint, which was shipped with all Macintosh computers, ensured that scholars
would soon begin tinkering with digitization of emblem pictures as well as texts.
One of the earliest scanners was the Apple Thunderscan, which was released in 1986.
single-line scanner replaced the ribbon cartridge in a dot-matrix Apple ImageWriter printer and
enabled the scanning of images from single sheets of paper that could be fed through the print-
er mechanism.While the Thunderscan worked as advertised,
it had many drawbacks. It was very
slow, since it scanned a single horizontal line of pixels with each pass: to scan even a small image
could thus take several minutes. Its resolution was very poor by today's standards, since it was
directly based on that of the ImageWriter and the Macintosh screen. It was a "one-bit" scanner:
in other words, it produced only black and white images rather than grayscale or color; while this
drawback was by no means disastrous where emblem books were concerned, it must be said that
the images were relatively crude as a result. Because of the mechanical motion of the printer
mechanism, the scanner had a tendency to go slightly out of alignment as it progressed from top
to bottom of an image. Finally, it could scan only images on single sheets of paper: in practice,
this meant scanning photographs or, more commonly, photocopies.
These drawbacks were soon remedied with the widespread availability of the first flatbed scan-
ners, which offered far greater speed, much better resolution, and the ability to produce grayscale
and then color images.This came at a steep price, however: while my Thunderscan had cost me
about $150 when I bought it in 1987, my first HP ScanJet cost nearly ten times as much a few
years later. Coupled with the steep price differential which in those days attached to Macintosh
computers (my first Mac Plus, purchased in 1986, cost over $4,000) compared to PC clones run-
ning a command-line operating system such as CP-M or DOS, this made emblem digitization
almost inaccessible for anyone without a substantial amount of research funding.
The release of Bill Atkinson's free HyperCard program in 1987 made available for the first time
a software solution that offered emblem scholars the promise not only to digitize isolated images,
but also to link them to one another and to their accompanying texts.
In other words, it made
possible the creation of the first widely available hypermedia databases, including the "Macintosh
During the final years of this first phase of emblem digitization, the use of
technologies such as HyperCard enabled the field of emblem digitization to develop to a point
where visual databases acquired a degree of maturity sufficient to persuade many emblem schol-
ars, even those without an intrinsic interest in digitization, of the importance of this emerging
Despite this progress, the closing years of the first phase, between 1987 and 1993, made clear
the fact that the current state of emblem digitization was unsatisfactory for a number of reasons.
The most important of these concerned the proliferation of competing proprietary standards in
computer hardware and software. HyperCard, for example, was available only for Macintosh com-
puters. The proliferation of competing graphics standards posed a serious problem. While the
CCITT and ISO working groups on bit-mapped photographic encoding had begun meeting
jointly as early as 1985 as the "Joint Photographic Experts' Group," the JPEG standard itself was
not published as ISO IS 10918-1 until 1994.
In the meantime, any number of different stan-
dards emerged, found favor, and disappeared, including the original MacPaint format, the PICT
format which existed in multiple versions, and the Graphics Interchange Format (GIF) developed
by Compuserve and introduced in 1987.
In the case of textual encoding, a similar situation applied.While the origins of the Standard
Generalized Markup Language, or SGML, can be traced as far back as the 1960s, SGML itself
became an ISO standard only in 1986.
Even after its adoption as an international standard, its
method of implementation was by no means a simple matter, and only the work of the Text
Encoding Initiative (TEI) from 1987 onward eventually produced a workable set of standards.
For early graphics formats includ-
ing MacPaint, see "From MacPaint
to JPEG: File Formats for Computer
The images in Graham (1991) and
Graham (1992), see footnote 1, were
produced with a Thunderscan.
Raggett/book4/ch02.html for a dis-
cussion that places HyperCard in the
context of HTML and the world-
See footnote 1, Graham (1991) and
See Robin Cover's history of
these developments at http://www.
DC_Emblemsbook_180204 19.02.2004 11:25 Uhr Seite 14