BMCR 1996.04.33

1996.4.33, Lange, Computer Aided Text-Reconstruction/Transcription

, , Computer aided text-reconstruction and transcription : CATT-manual. Tuebingen: J.C.B. Mohr (Paul Siebeck), 1993. xii, 160 pages : illustrations ; 24 cm. ISBN 9783161461491.

This book is not just for Holy Scrollers, although all its examples are taken from the Dead Sea Scrolls. Anyone engaged in the decipherment of ancient manuscripts and inscriptions might learn from its careful explanation of how images can be manipulated with the help of special programs to make the writing easier to read. In fact, all those who use visual materials in their study of the past could profit from the kind of programs described in the book, although its focus is clearly on written texts. There is no reason why an image of, e.g., a red-figured vase could not be subjected to the same kind of manipulation as a text to make the original design of the painting stand out. Unfortunately, the book is by now two years old and this means that it no longer reflects current technologies. But as is usually the case, those wholly innocent of imaging technology might profit from it nevertheless.

What the book offers is a fairly repetitive description of various programs (e.g., PhotoStyler) to manipulate images to be used with an IBM compatible computer. Lange describes a system with 16MB RAM and a 500 MB hard disk (a “medium configuration”), but larger systems are now well within our reach. The manipulations described include: magnification, analysis by histograms, creation of a negative, regulation of brightness and contrast, gamma correction, mapping, equalization, comparison and joining of fragments, relocation of mirror writing, comparison of characters and reconstruction of text gaps. Some programs now offer many more sophisticated manipulations, but the ones described are still basic. Magnification needs no comment; it is the computer equivalent of using a microscope. Analysis by histograms is more complicated. It allows one to manipulate highlights and shadows and thus to regulate brightness and contrast. The latter can also be done by regulating brightness and contrast separately. A negative is sometimes better to read for the human eye; it is the computer equivalent of a photostat. The gamma correction calibrates the monitor. Mapping and equalization increase the contrast and therefore the legibility of trouble spots. Comparing or joining fragments is fairly easy on a screen, but the images need to be in the same scale. This can be achieved by using the same dpi rate. Relocation of mirror writing is obviously useful for ink texts, but it might also be useful for joining an image of a squeeze of one part of an inscription with an image of the original or a photo. It will be more difficult to control the scale in the latter case. Comparison of characters from one part of a manuscript where they are more legible with characters from another part where they are faded, darkened or otherwise damaged might occasionally be helpful. Reconstruction of text gaps is only feasible with texts written in bookhands or with inscriptions.

What the book does not offer and might interest users is the following. It does not describe adequately how the images are acquired and how this affects what one can do with them. It simply assumes that 300 dpi grayscale images are made from black and white photographs with the help of a flatbed scanner. Other possibilities have now come within our reach, e.g. when 600 dpi color images are made from originals with a digital camera. 600 dpi is the suggested dpi rate for archival images of original manuscripts suggested by the Advanced Papyrological Information System, an undertaking of six American institutions (Columbia University, Duke University, Princeton University, The University of California at Berkeley, The University of Michigan and Yale University). This resolution is not strictly necessary to profit from the manipulation features of the programs described by Lange, but it captures more details. An example: on a typical 72 lpi screen a 300 dpi image can be displayed at about 4x magnification; a 600 dpi image can be displayed at about 8x magnification. Color images allow infinitely more manipulations, especially if they register infrared. Through multispectral imaging darkened passages in the Dead Sea Scrolls have been made visible that were intractable with the programs described in Lange’s book. Of course this requires access to originals, more storage space for the much larger files and equipment and programs that are still beyond the reach of ordinary mortals. Using originals eliminates a level of possible distortion introduced by using a print or a slide. A camera can be used for three-dimensional objects such as ostraca (potsherds) and inscriptions on stone. The increase in possibilities has made us aware of some of the uncertainties involved in digitizing ancient manuscripts and inscriptions. Not all cameras register infrared. To achieve “true” color a camera or scanner needs to be calibrated. Even so, digitizing a manuscript framed in glass will yield different results from digitizing it out of its frame. When one uses a camera different light sources influence the result differently. Using an underlight generally improves the contrast between otherwise black holes and ink and may also help in scanning squeezes of inscriptions. Using a sidelight generally improves the legibility of stylus tablets of any kind.

The internet falls entirely outside the scope of Lange’s book, but it is important to note the possible effects of sharing images from various collections worldwide. We could all draw on each others’ squeeze collections to check readings from inscriptions or fit papyrus fragments from one collection to those from another. Institutions that do not hold collections of ancient manuscripts or squeezes of inscriptions could still use images of those in other collections for educational or research purposes. The implication is that we all have to learn how to manipulate the images that are becoming increasingly available. For a substantial example see the Duke Papyrus Archive (http://scriptorium.lib.duke.edu/papyrus/).

The manuals accompanying the programs that are currently on the market (also for a Macintosh system) describe the various features to manipulate the images about as adequately as Lange’s book. But their main focus is on designing commercial advertisements rather than reproducing existing materials faithfully and legibly. The value of Lange’s book lies in the hundreds of illustrated examples it gives of using the manipulating features on ancient manuscripts. They invite us to use these features on similar manuscript materials at our disposal.

  • [1] The lateness of this review is the editor’s fault.