Dabire Collective

Arabic printing did not start in Europe, as is sometimes stated. In fact the Arabic script first appears in printed documents in Egypt, as early as the 10th century CE. Although these include passages from the Qur'an and other Islamic religious texts, it seems that they were not really designed to be read, nor were they generally intended to be artistic: some pieces have rather crude decoration calligraphy was beyond the capability of those who produced them. In fact, they were intended as amulets, to be rolled up and worn in a locket, to ward off evil, and it seems likely that they were printed from wooden or metal blocks, so that they could be produced cheaply, for sale to the illiterate poor. They were all single leaves, or two or three strips joined together at the edges. Longer texts were not attempted, as far as we know, and no books were produced in this way. The technique seems to have died out around the 14th or 15th century. What these pieces show, I think, is that the idea that texts produced by Arabs or Muslims always used well-formed calligraphy, whereas those produced by European non-Muslims did not, while not entirely untrue, is an over-simplification. This may become further apparent in what follows.

Arabic typography and printed book publishing did originate in Europe, in the 16th century, and developed there over the next three centuries, before printing was reintroduced to the Arab world. This birth and early development, however, were by no means painless, and I intend to show, among other things, how far the early producers of Arabic printed texts fell short, not only of calligraphy, but sometimes even of basic legibility. However, as we have seen, there was a precedent for this among the Arabs themselves.

The first printed book to contain Arabic script in any form was Breydenbach's famous account of his journey to Palestine, first published in Latin in Mainz in 1486. It included a table of the Arabic alphabet [see below], depicted, and probably also engraved, along with topographical and other illustrations, by the artist and printer Erhard Reuwich, who had accompanied Breydenbach on his travels. Reuwich presents the letters individually, in approximations of their isolated forms, and no attempt is made to indicate the variations in the linked forms - except possibly in the case of kaf, where we seem to have the initial version, followed by what may be the final or isolated version, which is, however, rendered as an additional lam in the romanised names in the table. The ligature lam-alif is given as a separate character below. Other peculiarities are the misleading diacritical marks on the letter sin, and the doubling of the letter ha' - "hehe" - the second one may possibly be intended to represent the final form. Finally, the box in the lower left-hand corner seems to contain an attempt at a whole word: wa-'l-salam or fa-'l-salam although some letters seem to be missing, and it is barely legible.

Reuwich was a highly regarded artist, and is reckoned to have been the first to have made accurate illustrations for a printed book. We can only speculate as to the written source of these Arabic letters: maybe a local scribe, perhaps an Arab Christian, wrote out the alphabet on request, and recited the names of the letters. While Reuwich's attempt to reproduce it could have been worse - at least most of the letters are recognisable - it is apparent that he saw them through a kind of mediæval European gothic distorting lens. The alif, for instance, is rendered rather in the manner of a gothic capital I, and most of the other letters have angularities, serifs and appendages characteristic of contemporary German and other European manuscript hands and early type-faces. The aesthetics of Arabic calligraphy were clearly completely foreign and incomprehensible to Reuwich and his contemporaries. We can hardly blame them for this, but quite a long time - at least 100 years - were to elapse before this barrier began to be broken down, and printed Arabic appeared in Europe that began to approximate even to orthographic, let alone calligraphic norms.
But in any case, I doubt whether the presentation of the alphabet was seriously intended to help travellers or to convey information to scholars in Europe. I suspect that it was probably included for its entertainment value, as another exotic image, along with the depictions of human groups in Palestine.
 
Breydenbach's book was reprinted several times, and translations appeared in a number of European vernaculars, over the ensuing century. The Arabic alphabet appeared in all of them. In some cases further errors were introduced. This alphabet was still being copied in European philological works as late as the mid 17th century.

The first printed book to contain recognisable words in the Arabic script appeared at the end of the 15th century [above picture]. This book, the Hypnerotomachia Poliphili, printed by Aldus Manutius in Venice in 1499, was very much a product of the Italian Renaissance imagination, being an extravagant essay in magic, classical erudition and eroticism. It has many engraved illustrations, two of which incorporate words in the Arabic script. While some attempt has been made here to suggest calligraphy, probably based, as the Italian scholar Angelo Piemontese has suggested, on Islamic monumental epigraphy, the result, as you see, is far from authentic. In the word majd, for instance, the dot below the jeem is misplaced under the meem, and the dal is disproportionately high, so that the word could almost be misread as "bi-hall". In al-mahabba, an unwanted dot has intruded under the ha', making it look like a jeem, and the ta' marbuta is undotted. The illustrator's ignorance of Arabic is confirmed by the error in matching the words majd Allah and majd al-dunya with their Hebrew, Greek and Latin equivalents. Nevertheless, Piemontese has convincingly established a relationship between these woodcut inscriptions and the Arabic handwriting in contemporary Italian manuscripts written by Renaissance scholars.

The Latin inscriptions in other illustrations in the Hypnerotomachia were produced by the use of letterpress inside the woodcut images. For the Arabic, however, no such use of types was attempted, even though the printer, Aldo Manuzio, was a master-typographer with extensive experience in Greek and its complex ligatures. This shows how far Arabic printing had yet to go before it could match what was already being achieved in other languages of scholarship.

Another printed Arabic alphabet [above picture] appeared soon after the beginning of the 16th century, this time in Spain, in a primer entitled Arte para ligeramente saber la lengua araviga, written in Spanish by the monk Pedro de Alcalá and printed by Juan Varela in Granada in 1505. This was clearly superior to that of Reuwich, and done by someone with knowledge of the language. It was intended to serve a practical purpose: training Christian priests and others involved in converting the surviving Muslim population of Andalusia to Christianity. The script naturally shows a Maghribi-Andalusian influence, both in the shapes of the letters, and in the forms of fā' and qāf, with just one dot under and over respectively. In this table, medial forms of some letters are shown separately, e.g. 'ayn and ghayn. But the author of this volume was unable to proceed beyond this woodcut alphabet, and a few woodcut letters inserted elsewhere. Otherwise, in this book, more substantial Arabic vocabulary lists, and renderings of liturgical texts, are given entirely in romanisation. Arabic typography had not yet arrived.

It did arrive, less than ten years later. But it was a long time before it was sufficiently developed to provide an adequate basis for the normal printing of Arabic texts. Even then, its expense was often prohibitive. So other techniques continued to be used as a necessary expedient when Arabic script was required. The most common method in the earlier period was the woodcut, which involved cutting away the surface of a block of wood to leave the printed design standing out in relief; the block could then be used either by itself, or incorporated into a page of typeset text. That was the method used in the examples already discussed.

Another technique was engraving on to wood or copper plates, and then allowing ink to fill the incised design before the plate was applied to the paper. Both these techniques were originally devised to print pictorial illustrations in European books; but they could be, and were, also used to reproduce the Arabic script. In some cases, this was simply because Arabic types were not available or too expensive; but sometimes, I think, it was because Arabic letters and words were regarded not primarily as texts, but as exotic appendages, more akin to fanciful pictorial illustrations than to the European-language texts which they accompanied.

Examples include the first Arabic to be printed in England, in a philological work, Robert Wakefield's Oratio de laudibus & utilitate trium linguarum Arabicæ Chaldaicæ & Hebraicæ, which appeared in London in about 1528. The author complained about the lack of adequate types for his purpose, but the printer, Wynkyn de Worde, managed to cut a few words, presumably provided by the author, on to wood-blocks and insert them into the text. He was a pupil of the famous William Caxton, the first English printer, and used his initials in his printer's device which appeared at the end of his books. In this one [picture above], he has incorporated the besmala below his name. The style is quite crude, and he has left gaps between joined letters: maybe this was because of the difficulty of achieving cursiveness with the tools which he had available, or maybe just because he did not know how to do it. He also seems to have had an incorrect notion of the final form of the mim. But it is still startling to see the basmala used in this way in an early-16th-century English book.

Another use of the basmala, appears in a tract by Bartul Đurđević, Croatian Latinist of the 16th century, printed in Vienna in 1548 for use in Christian-Muslim polemical dialogue [above picture]. This is much superior to the earlier London one, and can perhaps be considered as at least an approach to calligraphy. But some of the letters still have some unauthentic peculiarities, such as the final mim, and the excessive upstroke in the hā' of Allāh. This inscription seems to have been done by the intaglio method, whereby the script was incised into the wood and left ink-free, while the background was inked - hence the striking white-on-black appearance.

In 1583 an entire Arabic book, part of the New Testament, was produced using woodcut text. It was printed in Heidelberg, where no Arabic types were available at the time, and represented a not very successful attempt to reproduce the calligraphy in the 13th-century Christian Arabic manuscript from which the text was taken, at that time in the Palatine Library in Heidelberg. The engraver was not sufficiently skilled to do this satisfactorily: the letter forms and strokes were clumsily executed, the relative proportions of thick and thin strokes, and horizontals and verticals, were not observed, and some letters were defective, for example the letter kāf, which in several places has lost its top-stroke. The editor Ruthger Spey intended to send this and other Biblical texts in Arabic for widespread distribution in the Arab world, as he explains in his preface. It is unlikely that many copies of this one were sent, not even among the Christians.

In the examples discussed so far, woodcuts were used mainly because Arabic types were not available. But even when and where they were available, woodcut and engraved renderings of the Arabic script often continued to be used. Sometimes this was because a larger size of lettering was required, usually for a display title. For example the Medici Press in Rome, which was famous for its Arabic types, used a woodcut title, executed in a calligraphic style and printed in red, in one of its late-16th-century editions. It was probably intended for sale in the Arab world, but the lettering, although impressive, is not quite convincing. More than a hundred years later, a typeset edition of the Qur'ān was published in Hamburg by Abraham Hinckelmann in 1694. This too used large woodcut lettering for its Arabic title. The style here looks somewhat grotesque, although I am told that something similar can be found in some Arabic manuscripts. But as well as being unacceptable to Muslim Arabs both theologically and grammatically, it is also deficient orthographically, as it lacks the dots under the yā' in al-Islāmīya.

As well as the need for larger sizes of lettering than were available typographically, there was sometimes a need for smaller sizes. In the small-format grammar of Christian RAVIS printed in London in 1649 (A generall grammer for the Hbrew [sic], Samaritan, Calde, Syriac, Arabic, and Ethiopic tongue), the elaborate tables of the verb forms [picture above] required something neater and more delicate than the type-faces then available, as well as being fully vocalised. This was achieved by the use of copper engraving, and the result seems quite successful.

Engraving was also sometimes used when Arabic script needed to be incorporated into illustrations. A magnificent copper-engraved portrait of Anthimos, Patriarch of Jerusalem [picture below], which forms the frontispiece of his edition of the Psalms which he had printed in Vienna in 1792, uses small neatly engraved Arabic script for the laudatory verses which appear beneath, as well as for the names and titles in the borders.


Another use was for representing Arabic inscriptions or seals. Some woodcut Persian nasta`līq titles from the seal of the Mughal emperor of India appeared in a popular compendium of travel literature published in early 17th-century England (Purchas his pilgrimes, London: William Stansby, 1625). They are not quite authentic, but are legible.

The standard of the Ottoman Grand Vezir Kara Mustafa, captured after the siege of Vienna in 1683, is reproduced in a printed broadside of the time, complete with its Arabic woven inscription, rendered in woodcut. It seems to be a reasonably authentic rendering, at least of the central basmalas.

It also needs to be mentioned that there was some interest, in early modern Europe, in Arabic script and calligraphy as a subject in itself. Writing books, compiled by writing masters, containing analyses of the shapes and proportions of letters, and setting out calligraphic models to be followed by scribes, were a popular genre of publication, especially in 16th-century France and Italy. Some of them included Arabic alphabets, although usually more as comparative curiosities than as models for emulation. An example is the Champ Fleury (Paris 1529) by Geofroy Tory, one of the greatest calligraphers and lettering designers of early modern France. His Arabic alphabet, which he claimed to have copied from his Italian contemporary Sigismondo Fanti, has some of the Gothic-looking angularity seen in Reuwich's attempt back in 1486, but with a more humanistic sense of proportion. It remains limited, however, to the isolated forms, and there are some notable inaccuracies, especially in the fā' and qāf.  In another 16th-century writing-book, by the Italian master Giovanni Antonio Tagliente (La vera arte delo excellente scrivere, Venice, 1524), the letters are more elegantly proportioned, but it still contain anomalies.

Attempts at more elaborate Arabic calligraphy are sometimes found in printed books. In a type-set edition of an Arabic grammar (‘Izz al-Dīn al-ZANJĀNĪ. Kitāb al-Taṣrīf / Liber Tasriphi) published at the Medici Press in Rome in 1610, a quite remarkable version of the basmala can be seen [picture above], engraved on a wood-block, with elaborate loops and curls. I do not know whether it corresponds to any authentic manuscript exemplar, but it is undeniably impressive.

I have dwelt at some length on this phenomenon of non-typographic printed Arabic in early modern Europe, both because it seems to have received less attention than it merits, and because it was a forerunner in some respects of a later method of printing texts in the Arabic script which became popular in some parts of the Muslim world: lithography.   This permitted a more holistic approach to the creation of texts for printing, by combining manuscript techniques with mass production, and avoiding the segmentation of the script through the use of movable types. However, in the 20th century typography reasserted itself as the normal means of producing Arabic-script texts for printing in most areas, including Iran, and the design of movable types became once again paramount. So it is worth looking briefly now at the origins and early development of Arabic typography in Europe.

As is well known, the first Arabic printed book was the Melkite prayer-book produced on the orders, and at the expense of the Pope at Fano in Italy (or possibly at Venice) in 1514. Here again we find woodcut calligraphy in the opening Christian basmala, but the main text is composed from movable types. It can be seen immediately that that is what they are, because of the small gaps between joined letters. We know almost nothing about who designed and cut the types, nor about what he may have used as script models, although it seems likely that he would have worked from a Lebanese or Syrian manuscript of the text. The result is crude but mostly legible.

The second Arabic type-face to be made, as far as we know, was that used in the polyglot Psalter published in Genoa two years later. This shows some influence of the Maghribi style of script, for instance in the shape of the final yā'. Perhaps this was because of Genoa's trading links with North Africa.

The third Arabic printed book is something quite different, both in its choice of text and in the nature of its type-face. It is the first printed text of the Qur'ān, produced in Venice in the 1530s, apparently as an export commodity to the Muslim world [above picture]. The type-face is definitely superior to its two predecessors, both aesthetically and in terms of legibility. But the text and typography of this Venice Qur'ān have been studied in detail by Maurice Borrmans, and he has pointed out a number of serious deficiencies. The letter dāl seems to be replaced by the dotted dhāl throughout, although as far as can be seen the real dhāl is a little different, with the dot slightly offset to the right. The thā' is also reduced to a tā', with two dots, and the alif maqṣūra is erroneously given two subscript dots like a terminal yā'. But the most serious defect, which really prevents this version being read correctly as a Qur'anic text, is the lack of true vocalisation = ḥarakāt. At first glance, it appears to be vocalised, but in fact only the fatḥa is present, and the kasra and ḍamma are entirely missing – and, worse still, they have been erroneously replaced by the fatḥa in many places. These defects, together with numerous textual errors, would have made it quite unacceptable to Muslims, if copies were ever sent to the Muslim world, which seems unlikely. Borrmans speculates that in fact multiple copies were never printed, and that the sole surviving example is in effect an uncorrected proof for an abortive edition.

Although in this Qur'ān some distinct sorts are discernible, for instance an initial yā' and a medial hā', in other places complete words or groups of letters seem to have been cast as units. This technique, known as logography, was sometimes used later for Arabic  -- an example, rather crude, can be found in John Selden's Titles of honor (London: William Stansby, 1614). It was also envisaged in the early 19th century as a solution to the problems of movable type for the more difficult Arabic script styles, such as nasta`līq. But by its very nature, it undermined the practical and economic advantages of using movable types in the first place: if typographic sorts are used for whole words, a huge fount, comprising in effect a dictionary of the language, will be needed.

In the 16th century, Italy remained the preeminent country in the field of Arabic printing.  In Rome, the Jesuit College printed a number of Christian texts from the 1660s onwards. Their type-face was rather feeble and anaemic, with somewhat stunted letter-forms –e.g. the kāf, with its tiny top-stroke, added almost as if it were a diacritic.

The great breakthrough came in the 1580s and -90s, when the French master-typographer Robert Granjon arrived in Rome, and designed and cut a series of new Arabic founts, in four sizes, which were used mainly by the Typographia Medicea [see example above, from the illustrated Gospels (Al-Injīl al-Muqaddas) of 1590-91]. These were clearly based on calligraphic models, probably provided by the scholar Gianbattista Raimondi, the director of the press, who had assembled a collection of Arabic manuscripts brought from the Arab world. Granjon showed great skill in creating faithful renderings of the Arabic letters, and, perhaps even more important, he realised the need to incorporate a large number of ligatures in his founts if they were to approach calligraphic standards. They achieved a level of excellence seldom matched in the ensuing centuries, and were revived and reused in 19th century scholarly publications in France and Italy.

They also inspired the design of a number of other Arabic founts in 17th-and  18th-century Europe. The French scholar-diplomat François Savary de Brèves commissioned a beautiful type-face, based on Maronite manuscripts, which was first used in Rome in 1613, and later transferred to Paris, where it appears in scholarly and religious publications until the end of the 17th century. It too was revived at the end of the 18th century, and was among the types used at the press of the French expedition in Egypt, 1798-1801.

Other type designers also followed Granjon, such as the Dutch scholars Raphelengius and Erpenius [his bible extract, Leiden 1615, is shown above], the German Peter Kirsten and the Englishman William Caslon. Each of these, however, imposed distortions and limitations on their Arabic type designs, partly because of their own failure to appreciate the subtleties of Arabic calligraphy, and partly because of the need to economise with ligatures and other refinements in order to create practical and affordable founts. The same was true of the many other Arabic typographers in different areas of Europe down to the 20th century.

The first Arabic types used in Muslim and Arab countries, in 18th-century Turkey, Syria and Lebanon, suffered also from this legacy, since their types too were to some extent modelled on their European forerunners. Only in the 19th century did this begin to change, and only in the 20th and 21st centuries have new typographic norms for Arabic entered the arena.

About the author
Dr. Geoffrey Roper is a bibliographical consultant and print historian. He was head of the Islamic Bibliography Unit at Cambridge University Library, 1982-2003, and Editor of Index Islamicus and of the World Survey of Islamic Manuscripts. He has written and lectured on the history of printing and publishing in the Muslim world, and curated an exhibition on the subject at the Gutenberg Museum in Mainz. He is an Associate Editor of the Oxford Companion to the Book and of the forthcoming Handbook of the history of printing in Europe and around the Mediterranean (Vienna 2014).