Experimenting with Maghribī scripts: tools and techniques

On November 7th, 2023 we had the pleasure of welcoming Umberto Bongianino, Departmental Lecturer in Islamic Art and Architecture at the Khalili Research Centre, University of Oxford, for a session on praxis. These sessions are meant to be exploratory collaborations between researchers and calligraphers, in order to experiment and try out new hypotheses, following an established protocol developed by the CallFront Project. In this case, this was a preliminary session ahead of our work with the Moroccan master calligrapher Hamidi Bel’aid, who will be collaborating with us in the upcoming future.

We structured this session around a series of questions posed by Umberto Bongianino concerning the instruments used to write the Maghribī mabsūṭ (“monumental”) style, as well as the techniques used for Maghribī thuluth chrysography.  The final part of the session broached questions regarding letter forms in epigraphic Maghribī thuluth, as well as the possible methods used for the transposition of models.

Fig. 1. Printed model from the Marrakesh Quran ca. 1220, the experiment sheet and different types of qalams.

            For the experiments on Maghribī mabsūṭ, we used three specimens printed to scale: two Qurʾāns possibly from Marrakesh, dating from ca. 1220[1] and ca. 1230 (the “Pink Qurʾān”),[2] and a Qurʾān probably from Fes, dated 705 H./1306.[3] Our main objective was to recreate the calamus which would have been used for these calligraphic styles, a monumental variant of Maghribī round scripts used especially, but not exclusively, in Quranic manuscripts. In order to do this, we first selected thick 160 gr drawing paper, with no sizing but polished, resembling the weight and texture of the Pink Qurʾān.  For the ink, although the corrosive effects on the pages of the original specimens indicate the employment of metallo-gallic inks (ḥibr), we used a black soot ink (midād), easier to procure and which behaved similarly to ḥibr in terms of consistency.

Medieval Maghribī mabsūṭ is a large script, characterised by a thick ductus and very little contrast between thick and thin strokes, especially in the articulations of letters, generating an overall “bold” appearance. To reproduce the calamus which would have been used for this script we chose to work with cane, still used by calligraphers in modern Morocco, and which were possibly also employed in medieval times by Maghribī copyists. We used Spanish cane (arundo donax) sourced from Valencia, of approximately 15mm in diameter, which is softer than the hard, thinner reeds (phragmites communis) used in the Middle East.  Following the models known for Maghribī script qalams,[4] we split this cane down the middle lengthwise, producing a width of about 12mm to be held comfortably between the fingers. For the nib, contrary to smaller Maghribī variants written with a narrow, rounded nib, the thicker ductus of Maghribī mabsūṭ called for a width of approximately 4mm, cut to scale using the printed models as guidelines. We cut the tip of the nib at a right angle and bevelled the edges in order to reproduce the blunt, rounded effect of the writing (fig. 2). 

Once this initial calamus was ready, we started by reproducing certain words from the models, including vertical strokes, bowls, diagonal strokes, loops, and shorter tails (fig. 1). We soon realized that while the nib reproduced certain initial strokes accurately, its thickness was not sufficient, creating too much nuance and thin transitions between strokes.  So, in addition to the increased nib width, an increased thickness of the nib was necessary to recreate not only the “bold” strokes but also the thick transitions and heavy letter articulations. Since we did not have a cane with enough density around the circumference, we experimented with a piece of soft balsa wood, cut to imitate the shape, width, and thickness of a cane qalam (fig.3). This thicker nib reproduced the strokes, liaisons, and bowls perfectly, which lead us to conclude that a cane with a larger diameter and hence with a thicker cross-section would have been the ideal choice for this type of script. This could also explain the use of canes split lengthwise in Maghribī calligraphy, a choice which had intrigued us from the outset:  a thicker, denser cane would imply a larger diameter, too large to manipulate between the fingers of the calligrapher, while a section of cane cut lengthwise would allow for more ergonomic comfort when writing (fig.4).   

Fig. 2 – Split cane with flat nib, 4 mm wide, with rounded edges / Fig. 3 – Side thickness of cane calamus (left) and balsa wood reproduction (right). / Fig. 4 – Cane diameters in relation to thickness and density.

Our next experiments concentrated on Maghribī thuluth chrysography, a calligraphic style mostly used in the colophons of illuminated manuscripts, and we worked with specimens from North Africa, dating from ca. 1175,[5] 599H./1203,[6] ca. 1310,[7] and 1066 H./1652-3.[8] Our main question revolved around the shape of certain characters and the overall process of production. For the gold we followed the technique which is still used for eastern scripts. A booklet of 22 K gold leaf was hand ground and made into a solid powder, following the traditional steps involved in gold ink making (fig.5). The solidified gold powder was then emulsified with a drop of gelatin,[9] diluted in water and applied with a brush to the nib of the qalam. We thus proceeded to the reproduction of a few words, on the same 130 gr paper (fig. 6). Once the letters were dry, the gold was burnished (fig. 7) and then outlined (taḥrīr) in black with a very fine brush.[10] The long time and great precision needed for the outline made us question whether this stage would have been produced by the calligrapher or a by second scribe specialized in taḥrīr, especially in those Maghribī manuscripts that employ chrysography throughout, not just in their colophons.

Fig. 5 – Steps in gold ink preparation, from left to right, grinding the gold leaf by hand with a few drops of gum Arabic until the particles are pulverized, then filtered and washed with water.

We remarked that the effect of the black outline on the letters deforms certain shapes, such as the letter alif, giving the impression of a continuous upwards stroke in the place of a serif attached to a downward stroke shaft. The black outline was also used to accentuate certain looped characters, such as the final lām-alifs and hāʾs, and to play with the interweaving of strokes, adding a three-dimensional effect to the calligraphy. Another interesting feature of Maghribī chrysography is that the heads or tips of most letters are highlighted in red, which adds a distinct character to the writing and produces a more “painted” effect rather than written, accentuating the calligraphy’s three-dimensionality.

Fig.6 – Reproduction of a few words with gold ink and a qalam after one of the models.
Fig. 7 – Burnishing the letters before the outline.

            The final part of our session broached on questions posed by medieval Maghribī epigraphy, which displays a series of styles also belonging to the family of Maghribī thuluth. Our questions were related to the preparatory drawings of inscribed friezes, whether they were carried out directly on the hard material, or first on paper and then transferred to the surface to be carved or incised. A recent article on the gate of the Marinid ribāṭ of Shālla, Rabat, discusses a preparatory drawing painted in red ochre which was discovered on the wall of the ribāṭ, which lead us to question whether they would have been produced with a brush or some sort of soft qalam/writing implement.[11] A few rough experiments reproducing the same design with different tools revealed that while the main letter forms, ascenders, counters, and ligatures could be easily reproduced with both a brush or a calamus, certain sections, such as the top initial flatter head of the ascenders, was difficult to reproduce with a round brush (fig.9). It is possible that the contours of letters were hand drawn to exaggerate or stylize certain letters, creating a particular aesthetic for this epigraphic variant of Maghribī thuluth. More research into contemporary techniques for the transposition of models and the drawing of certain characters by living artisans may shed more light on how these letter forms were exactly produced. 

Fig. 8 – Umberto Bongianino and Nuria Garcia Masip at work.
Fig. 9 – Experiments with epigraphy models.

Nuria Garcia Masip & Vincent Thérouin


[1] Istanbul, Topkapı Palace Library, ms. R. 21, f. 227b.

[2] Marrakesh, Ibn Yūsuf Library, ms. 431/6.

[3] Bavarian State Library, Cod. arab. 2, ff. 58b-59a.

[4] Umberto Bongianino, The Manuscript Tradition of the Islamic West (Edinburgh: Edinburgh University Press, 2022) 38-39; François Déroche, Islamic Codicology: An Introduction to the Study of Manuscripts in Arabic Script (London Al-Furqan Islamic Heritage Foundation, 2005) 104-105.

[5] Fes, Qarawiyyīn Library, ms. 181, pp. 6–7.

[6] Istanbul, Topkapı Palace library, ms. R. 33, ff. 136b–137a.

[7] Rabat, National Library of the Kingdom of Morocco, ms. K 2949.

[8] Doha, Museum of Islamic Art, MS.1999.11.

[9] The gelatin, traditionally fish glue or very diluted gum Arabic, acts as a sort of binder so the gold does not rub off once it is dry. Medieval handbooks cite similar mediums, such as fish glue, snail glue or rice starch: see, for instance,  Martin Levey, “Mediaeval Arabic Bookmaking and Its Relation to Early Chemistry and Pharmacology”, in Transactions of the American Philosophical Society, 52 (4), 1962, 1–79: 9, 37. Whatever the medium used, it is important that the dilution is exact to avoid tarnishing the gold.

[10] We used a very fine Kolinsky sable brush, size 0000.

[11] Patrice Cressier et alii, “Un cas unique d’épure d’architecture en Occident islamique. La représentation de l’arc et du décor de la grande porte mérinide de Šālla (Rabat)”, in Arqueología de la Arquitectura, 18, 2021, 1-33. 


OpenEdition vous propose de citer ce billet de la manière suivante :
callfront (21 mars 2024). Experimenting with Maghribī scripts: tools and techniques. CallFront. Consulté le 25 juillet 2024 à l’adresse https://doi.org/10.58079/w25n


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search