The Arabic Papyrology Database


The paleography angles tool is based on two different research projects. The first is Eva Youssef-Grob‘s project on script angles, the results of which she presented in her 2013 article: Grob, Eva Mira. 2013. A catalogue of dating criteria for undated Arabic papyri with ‘cursive’ features. In: Documents et Histoire. Islam, VIIe-XVIe s. Actes des journées d’Etudes musée du Louvre/EPHE, mai 2008, 115–136. Ecole pratique des hautes études, Sciences historiques et philologiques II, Hautes Etudes Orientales-Moyen et Proche-Orient, 5/51. Genève: Droz). The second is a similar project by Sonego (publication in process). The two projects differ by their corpus: Grob evaluated some 700 dated documents from the first five islamic centuries in Egypt (97 before 800 CE, 290 from the 9th century, 168 from the 10th, 119 from the 11th and 56 even younger). Sonego measured some 80 documents from the same place and period, of which the majority is also present in Grob‘s corpus, and most have a typological mark-up. For both Grob and Sonego the condition was that the documents had to be dated by non-paleographical evidence, like explicit dating, archeological context or affiliation with a „family archive“. This means that letters are poorly represented, and with regard to the typological markup, there are more A2. and A3. documents than A6. or B. documents. Grob offers categoric data („low“, „intermediate“, „strong“), while Sonego gives a numeric value which is based on at least four measurements per document. While Grob did her measurements with a protractor, Sonego has been drawing coloured lines into digital images. Johannes Thomann programmed a Java application to extract the angles from the svg files. The images can be viewed under „ANGLE“ in a new tab, where available.

Slants and angles

▬ Orientation of overall writing line: General orientation is measured in relation to the margins or the fiber, in most cases identical to the margin of the photograph.

▬ Absolute slant: The hastae. Absolute slant is measured in relation to margins or fiber.

▬ Baseline: The baseline of words is defined as the straight line that goes through the lowest point of the joins between letters. It is only drawn as an auxiliary line for „hanging“ and „script angle“.

▬ Hanging: Hanging means the acute angle at the intersection of baseline and overall writing line.

▬ Script angle: Script angle means the obtuse angle at the intersection of baseline and hastae.

▬ Relative slant: Relative slant is the upper left-hand angle at the intersection of writing line and hastae. This angle is not visualized in the search results.

P.Vente 2 (2[6]1 AH / 875 CE): raḍiya-hā wa-tasāhala bi-hā

Search hint

The data will not give all values for all documents. Searching for script angle or relative slant, e.g., will give you less results than a search for absolute slant. We recommend to try different options and not to narrow your search by choosing more than two parameters. Geometrical consistence should also be considered, e.g. a text with an acute script angle is unlikely to have a slant to the right, unless the general orientation is conspicuously rising. Relative slant is not visualized in Sonego‘s corpus, but it can be calculated from general orientation and absolute slant for documents with preserved margins. In case the margins are lost, it can still be inferred from hanging and script angle, as the sum of angles in a triangle equals 180°. In documents with zero hanging, it will appear as script angle. In order to simplify the search routine, numeric values have been converted to categoric ones for the documents that Grob did not measure.

How to measure a document in order to date it

We used inkscape, a powerful vector graphics editor published under a GNU general public license. Thomann programmed an application to extract the coordinates of paths with a given colour, and to calculate slants and angles from the coordinates. This software has also been published under a public license. However, any graphics editor like Libre Office Drawing will indicate the slant of the line you are creating somewhere in the lower margin. Some hints for drawing: There is often a remarkable difference between alifs drawn from top to bottom and those drawn from bottom to top (this is what makes cursive hands „chirodictic“), so you should have a look at both of these. Drawing a straight slant over a curved alif is often a matter of style, I would suggest to play around and then use an average value. I defined the baseline as the straight line that goes through the lowest point of the joins, but letterblocks (graphic units) with two joins are not as frequent as one might expect. Short elements like a sīn or the base of an ʿayn are often more inclined than the word in its entirety, or single words are more inclined than groups of neighbouring words. I tend to measure the longer, flatter units in these cases. The most difficult slant in my opinion is general orientation, especially finding the left end in a curved line. I used to draw the line from left to right. The right starting point of a line is not the first letter, but the end of the first word.