Ancient writing systems in the Mediterranean

A critical guide to electronic resources

Ancient South Arabian

- 11th/10th c. BC - 6th c. AD

edited by: Alessia Prioletta (translation revised by Melanie Rockenhaus)

  • Introduction
  • Index
  • Further information

Example of inscription in monumental script and archaic style

In Southern Arabia, during the pre-Islamic period, two types of consonantal alphabetic writing were in effect: the monumental and the minuscule alphabets. The monumental script, also called capital, is regular and geometric and was engraved on supports designed to last and to be publicly seen. The minuscule script, written in a more agile ductus that resembles our italics, was recorded on perishable material (wooden sticks and palm-leaf stalks) and was mainly intended for private use.
The earliest documents date back to the 11th-10th c. BC and are minuscule inscriptions. The monumental writing was formalized only around the 9th-8th c. BC, when the South Arabian kingdoms were formed, all of which shared both these writing systems, with minimal stylistic differences. The most recent South Arabian documents date back to the middle of the 6th c. AD.

Go to the online resources.

Online resources


Origin and formation of the South Arabian alphabet

At the beginning of the 2nd millennium BC, the alphabetic writing system was invented in the ancient Near East. A few centuries later, two great traditions developed from this, each of them with its own alphabetical order and shapes and names of letters: the Phoenician and Aramaic system (or Northwest Semitic) and the Arabian system (or South Semitic).

The South Arabian alphabet, from which the Ethiopian syllabary and also perhaps the Hasaitic alphabet derived, belongs to the Arabic tradition, together with the North Arabian alphabet. It is not known exactly in which region the proto-Arabian alphabet was formed, whose oldest documents are some South Arabian sticks that could date back to 11th-10th c. BC according to Carbon-14 dating. The area of its use could have extended across the region of southern Syria, Palestine and Sinai. According to a recent hypothesis, the area of the oasis of Taymāʾ, in northern Arabia, might have constituted a place of exchange for the spread of the alphabetic tradition in the rest of the Arabian Peninsula.

Despite a certain similarity with some characters of the Phoenician alphabet, the existence of different signs made it immediately clear to scholars that the South Arabian alphabet belonged to a different tradition. This was confirmed by the discovery of some abecedaries engraved on slabs, rock or stones found in Timnaʿ and Maʾrib (Yemen), and in al-ʿUla (Saudi Arabia), which had a different order than that of the Phoenician alphabet. In 1987, the scholar A. Lundin recognized in a tablet found at Beth Shemesh, in southern Palestine, an abecedary written according to the South Semitic alphabetical order. In 1990, a similar document was also found in the city of Ugarit.

South Arabian alphabet order and composition
The South Arabian alphabet consists of 29 signs, two more than the Ugaritic alphabet and seven more than the Phoenician alphabet. Compared to most of the other Arabian alphabets, the South Arabian alphabet contains the grapheme s3, which indicated the third non-emphatic sibilant, dating back to the Proto-Semitic phonetic repertoire:

h l ḥ m q w s2 r b t s1 k n ẖ ṣ s3 f ʾ ʿ ḍ g d ġ ṭ z ḏ y ṯ ẓ

To these characters the sign | can be added, which separates each word. In ancient times and up to the Christian era, the numeric ciphers were described by some letters, which constituted the initial consonant of the indicated number:
| (1), (abbreviation for ẖms1t “five”), ʿ (abbreviation for ʿs2rt “ten”), m (abbreviation for mʾt “hundred ”), a symbol designed as half of m (to indicate “fifty”) and finally ʾ (abbreviation for ʾlf “thousand”). 

As mentioned, the first attestations of South Arabian writing could be found on wooden sticks with minuscule texts. The first examples of monumental inscriptions date back to the 9th-8th c. BC.

Lundin 1987
Robin 1991-1993 a
Bordreuil e Pardee 1995
Macdonald 2000
Bron 2002
Robin 2008

Monumental script

The most significant documentation, as well as the most copious, coming from Southern Arabia consists of monumental inscriptions, which the Arab-Islamic tradition recorded under the name of musnad. The monumental script shows great regularity, with signs which are separated from each other, generally easy to read and close to our capitals.

South Arabian monumental writing was formalized originally in the region of al-Jawf and Sabaʼ, from which the most ancient documents originate, and where some of the letters reveal a certain hesitation in shape (s3, and in particular). In addition, in some centers of al-Jawf, such as as-Sawdāʼ, there were attempts to impose particular forms for some letters (such as the r with a double stroke or the m with three points), which failed to be established.

Ductus and evolution of the script
The South Arabian inscriptions are generally read from right to left. However, the most ancient documents, of which we have some very long examples, sometimes have a boustrophedon cursus, in which the direction changes on each line. Boustrophedon writing was particularly widespread in the kingdom of Sabaʼ, where it continued to be used at least until the beginning of the 3rd c. BC, while it was much less popular in the other South Arabian kingdoms.

The studies of J. Pirenne have been crucial for the analysis of the South Arabian monumental script. The letters can be divided according to shape (rectangular letters, rectangular with diacritics appendix, with stroke and appendix, with circles, dome, etc.), the proportion of the axis, technique and ornamentation.

Of the 29 letters of the alphabet, 11 have an axis of symmetry and only 8 do not. Pirenne showed that the graphic evolution is based primarily on the variation of the ratio between the height and width of the characters. During the archaic stage, this ratio was 1:2, while in the "classical" script, with its perfect harmony, the ratio became 1:3.

Pirenne also noticed an irreversible trend in the form of the strokes of the letters. In older texts, the strokes are straight, but they gradually thicken towards their free ends, which take the form of elongated triangles. Subsequently, the thickening occurs exclusively in the triangle at the free end, as in the classic print fonts.

Another irreversible evolution in the monumental script is the transition from right to acute angles, which is evident in the letters with domes or in the rectangular ones with diacritical appendix.

Finally, from the 4th c. AD, most of the monumental inscriptions were engraved with a technique similar to the champlevé, in which the stone around each character is excavated, so as to create a relief effect. This style shows a certain agility perhaps influenced by cursive writing, and it reaches progressively extreme examples in which the parts excavated are minimized. This confers a confusing effect on the text, in which the characters are sometimes hard to distinguish.

Pirenne 1956
Robin 1991-1993
Ryckmans 1991-1993

Minuscule script

Discovered and deciphered in the early 1970s, South Arabian minuscule writing is defined in this way to distinguish it from the monumental or capital writing. The definition of cursive script tends to be rejected by scholars both because the word contains the idea of hasty and rough handwriting and because it is already used to describe the graffiti carved on the rocks of pre-Islamic Arabia, which have no relation with the minuscule script. In addition, in modern times cursive writing indicates letters that are joined, which is not the case with the South Arabian minuscule script.

The Arab-Islamic tradition called this script zabūr (pl. zubur) and considered it as the cultural heritage of the people of pre-Islamic Yemen, and it was only inscribed on palm-leaf stalks. Wood is, in fact, the almost exclusive material of this script, and there are only very few examples of minuscule texts that are engraved on other materials, such as rock or bronze in particular.

Currently, around 250 texts inscribed in minuscule have been published, but there are hundreds of other specimens, preserved in some museums of Yemen, especially in Sana'a, but also in European collections such as those of Leiden and Munich.

The radiocarbon analysis performed on some specimens have led to date some of these sticks to the earliest phase of South Arabian history, in the late 2nd millennium BC, a phase that so far was only known archaeologically and whose only examples of writing were made up of few isolated letters on pottery.

Ductus and evolution of the script

The ductus is very agile, despite the hard support that is quite difficult to engrave. The track is inclined towards the right and the lower ends of the letters extend towards the left, to give an effect of a wake in the direction of trace of writing. As in the monumental script, a vertical line separates each word.

The style of the minuscule writing clearly evolved over time. J. Ryckmans first analysed the main innovations and highlighted different graphic styles. While in the archaic stage there is a clear similarity with the monumental writing, the following stages showed independent evolutions, in which the letters take a different form from the corresponding capital letters. There is a gradual incline of the lines of the characters with a loss of their geometric shape. The letters with a fork, frame or hook elements develop in a vertical line with termination oriented to the left, and a circular semi-closed element to its right. Some series of letters thus end up being very similar to each other, making it difficult to interpret the text, for example the letters ʿ, b, m, or ʾ, k and s1. Others, such as d or , took a form radically different from that of the archaic phase. The last stage of the minuscule writing has a very specific ductus, with markedly vertical and broken lines and the loss of specificity of some characters that ended up being almost identical.

In terms of orthography, the minuscule writing very early lost the grapheme , whose phoneme was rendered with the , probably as the result of a phonetic evolution in the spoken language.

Ryckmans et al. 1994
Stein 2010 a

The discovery of the documents and the deciphering of the alphabet

The first report on the presence of inscriptions in Southern Arabia dates back to C. Niebuhr, a Danish explorer member of an expedition organized in 1762 by King Frederick V of Denmark. In 1810, U. von Setzen went to Sana'a in search of inscriptions and sent a first facsimile of a few letters to Europe before he died mysteriously in Yemen. The first long inscription was seen in 1834 by S. B. Haines and J. R. Wellsted, British sailors of the ship Palinurus, who copied a text of 10 lines engraved on the rock of Ḥuṣn al-Ghurāb, east of Aden.
The deciphering of the South Arabian alphabet is the work of two German scholars, W. Gesenius and E. Rödiger. On the basis of the similarity with the characters of the Ethiopian syllabary and with the help of inaccurate copies of South Arabian characters present in Arabic manuscripts, in 1841 they arrived to the identification of two-thirds of the 29 South Arabian letters.
In 1870, when J. Halévy travelled to Yemen, there were two letters to still be deciphered correctly: the s3 (read z) and z (read ʾt), in addition to the uncertainty about the value of t and . European scholars arrived at the complete deciphering of the alphabet at the end of 1800, thanks to the many good documents available, not only facsimiles, but especially moulds and photographs, in addition to numerous inscriptions that physically arrived to the museums of Europe.
Only in 1970, almost 130 years after the discovery of the first South Arabian monumental inscription, were the first documents in minuscule script discovered during illegal excavations. This writing tradition was already known from Arab-Islamic sources and these exemplars consisted of two sticks, 12cm and 16cm long and with a diameter of 3cm, inscribed with a 14-line text in an unknown script. A collector brought the sticks, said to come from as-Sawdāʾ, as a loan to the scholar M. al-Ghūl, who presented them at scientific conferences and deciphered the first few characters. The complete deciphering of the characters was reached, however, only in 1977.  

Ryckmans et al. 1994
Robin 1997

The role of palaeography in South Arabian studies

The most important traditional tools for dating ancient documents have only limited application in South Arabian studies.

The use of the titling of the kings mentioned in the inscriptions by their name, affiliation and epithet, generally helps determine the genealogical lists and calculate the approximate duration of the reign of each king. Here it is penalized by limited and repetitive royal onomastics, which generate many homonyms and therefore make it difficult to place the various rulers in sequence.
Moreover, very little data are available to anchor the inscriptions to an absolute dating. There are few mentions of South Arabian rulers in external documents (Assyrian annals or Greco-Roman sources) and the South Arabian inscriptions rarely make reference to contemporary historical events involving other regions of the ancient Near East.

Given these limitations and also the largely formalized character of South Arabian writing (especially the monumental script) and its evident evolution, palaeography has always been a widely used method for historical reconstruction.

After the first studies dating back to the late 1800s and the first half of 1900, J. Pirenne first founded the criteria for the chronological evolution of the South Arabian monumental writing. Having an available material of about 500 pictures and 700 facsimiles, Pirenne first considered palaeography as a true science of observation, with the purpose of analysing, naming and classifying documents. The material was then reduced to a common size (1cm) to maintain proportions, and the letters, arranged by shape, were drawn on a grid sheet. The acute aesthetic sense of the scholar allowed her to recognize the key letters in the evolution and to identify exactly the elements that changed over time. This work led to the classification of the inscriptions in periods, which are further divided into styles.

However, although Pirenne's palaeographic results still remain valid today, the historical conclusions that she drew have proved completely wrong. Pirenne believed that South Arabian writing stemmed from the Phoenician script and was influenced by the Greek script and culture. She also thought it originated in the 5th c. BC and therefore that also the South Arabian kingdoms rose at the same time. The application of the palaeographic method on the inscriptions of the last historical period (from the 3rd to the 6th c. AD) likewise led the scholar to often wrong historical reconstructions.

A few years later, H. v. Wissmann developed a new palaeographic-chronological grid of South Arabian inscriptions according to which the first documents were dated to the 8th c. BC. The so-called "long chronology", as opposed to Pirenne's "short chronology", has now been definitely accepted.

Although the general impression is that script evolved in a fairly homogeneous way across Southern Arabia, it cannot be imagined that the graphic innovations were incorporated simultaneously and in the same way by the various South Arabian scribal schools. In addition, it may have happened that there were graphic delays or styles that were local and quite distant from those in vogue in the central schools, especially in some remote areas.

The palaeographic method can prove to be very useful and reliable especially when applied to the documentation of a single site. A good example is the corpus from the site of Raybūn, in Ḥaḍramawt, studied by S. Frantsouzoff. Even in this case, however, the application of palaeography has been reduced to a mathematical calculation of coefficients of proportions, in which the corners and the infinitesimal variations of the shape have ended up having more importance than other data in the dating of the texts.

A more comprehensive approach, which takes into account not only the palaeographic ductus, but also textual features such as grammar, vocabulary, formulary and content, has been applied by A. Avanzini for the subdivision of the documentation from Qatabān.

Recently, P. Stein, whose study starts from the few inscriptions that present useful data for a certain chronological basis, has outlined a new framework for the development of South Arabian writing, either monumental or minuscule. He highlighted four main styles with internal changes, and an evolution that occurs about every three centuries.

To date, there is still no agreement among scholars on either the terminological or substantial levels, about the main graphic styles of the South Arabian monumental script.

Pirenne 1956
Wissmann 1976
Ryckmans 1991-1993
Pirenne 1990
Ryckmans 2001
Avanzini 2004
Frantsouzoff 2005
Stein 2013

Macdonald 2015


Materials and supports

Ancient Southern Arabian was written on an extraordinary variety of materials and different supports. The choice of these is not casual, but is closely linked to the type of text, and, in the case of the artefacts, often to the function of these.

Stone is by far the most widely used material. It is preferred, of course, for inscriptions that are incised on supports that can be defined as "monumental", such as the blocks in architectural structures: defensive works (city walls), irrigation structures (dams and reservoirs), temples or private houses. The stone is often natural rocks, and hundreds of inscriptions were carved into the mountains and cliffs of various Yemeni regions. The rock inscriptions can commemorate the passage of an army as they can that of a single shepherd behind his flocks. At some rock sanctuaries, the South Arabians have entrusted the memory of war campaigns to rock, which provide historical data of the greatest importance for modern scholars. Also engraved on rocks there are important legal texts that govern the division of the surrounding fields, the flow of water from the canals, the boundary between two territories. This function is also fulfilled by big standing monumental stela affixed in open places such as markets and cultivated fields.

The second most common type of material is wood, which was inexpensive and available in large quantities at any agricultural settlement. Hundreds of wooden sticks or palm-leaf stalks have been found, inscribed with texts in minuscule script. The Arab-Islamic tradition calls the stems of palm ʿaṣīb (plural ʿuṣub) and it is said that parts of the Qur'an have been written on this type of support.

Inscriptions are then incised on a wide variety of artefacts, often made of precious stone, such as alabaster, a material highly sought after and loved for its beauty and luminescence. Inscriptions were also engraved on gold and silver (goldsmith, glyptic) or organic materials, such as coral and ivory.

The Southern Arabian culture has also left a large number of inscriptions on bronze artefacts. Inscriptions were also engraved on a wide variety of sculptures, both anthropomorphic and theriomorphic, as well as on furnishings, cult objects (oil lamps, incense burners, jewellery), measuring instruments (weights, scales and units of measure), pottery and supports.

The different types of ceramic containers, used for the transport and storage of various products and food, often bear engraving on the wall or on the edge, a short inscription that may indicate a special propitiatory formula, the name of the owner, the substance contained or the quantity of the product. In southern Arabia several ostraca inscribed with a proper name have been found, used to locate the graves of the dead.

The contexts of use of these objects, often of high artistic quality, are mainly three: religious, in the case of votive objects offered to the deity and often accompanied by a dedicatory inscription (such as statues or decorated slabs) or which were part of the cult furnishings (altars, incense burners, offering tables), domestic (in the case of many categories such as furnishings, containers, supports) and funerary (funerary stelae and objects of the tombs furnishings).
Not documented archaeologically, apart from rare exceptions, are materials such as papyrus, leather and parchment, but also cotton, linen, silk and animal bones, materials that were used extensively in other regions of the Near East and the Mediterranean.

Maraqten 2009
Stein 2010 b


The most commonly used writing technique is incision. On the rocks and on stone, as well as on wood or pottery, the text was chiselled.

Inscriptions on stone were either engraved with a single stroke or hewn in relief. This latter technique becomes the norm for the last period of South Arabian history (4th-6th c. AD) when, apart from the inscriptions engraved on the rock, those on stones are almost invariably carved in relief.

The minuscule inscriptions incised on wood have a fine stroke that can be uniform, or have threads and bolds. This implies an incision by means of a very sharp blade. So far, iron or sharp bronze styli have been found, as well as in wood and ending with a metal tip. Styli in ivory have also been found. In some texts, the incision detached to form a relief of a dark colour, due to the swelling (caused by moisture or salinity) of the original substance (charcoal or ink) that had been included in the trace to improve writing.

Two main techniques were used for writing on bronze consisting of systems that could create embossed or engraved inscriptions. The relief technique is used only in the case of lost wax (both hollow and full) casting on a positive, and consists in modelling the letters with the wax and then attaching them to the wall of the mould. In the case of objects inscribed with this technique, text and support were created at the same time and so their simultaneity is ensured.

Incised inscriptions could be obtained with two different techniques. The first, feasible only in the case of lost wax casting on a positive, consists of drawing the characters in the wax through a wide and deep stroke to produce a legible inscription. In the second case, the text is engraved in bronze with the graver after smoothing. In that case, of course, there is no certainty that the inscription was contemporary with the support.

The inscriptions engraved on bronze may have a double-stroke script, in which the shape of each character is made by a double parallel line. Another technique of engraving on bronze is that of a zigzagged stroke. A series of miniaturised alabaster containers, such as small vessels and plates, as well as normal pottery containers are often inscribed with a dots technique.

Writing with ink or paint is almost entirely absent, although Arab-Islamic sources state that this technique was widespread in Southern Arabia in the 6th c. AD.

ʿAlī ʿAqīl and Antonini 2007
Jandl 2009
Ryckmans 1978
Stein 2010

Literacy in South Arabia

The epigraphic corpus from South Arabia comprises about 12,000 inscriptions, written in both monumental (for the most part) and minuscule script.

The South Arabian documents relate mainly to issues of everyday life of society. The most common textual typology is that of the construction inscriptions commemorating the building of temples, dwellings, defensive or hydraulic structures, tombs, etc. There are several texts of a religious nature, dedications addressed to the gods to request a favour or as atonement. Also connected to the religious life are the inscriptions that regulate the cult ceremonies, oracles, proverbs and omens. Lastly, legal texts are also common, both those drawn up by the authorities, such as decrees and prohibitions, and those commissioned by private individuals and written in minuscule script, which regulate economic and legal issues.

The documents that are not connected to everyday life include commemorative inscriptions (especially of military and political enterprises) engraved on the rocks of the mountains, sometimes in inaccessible places. Rocks are also carved with hundreds of graffiti, left by people passing by and that generally contain only a single proper name.

Literary, historical or mythological epic is almost entirely absent. One exception is the so-called "Hymn of Qaniya", a poem in verse engraved on the rock and still not completely understood.

The extraordinary amount of either official or unofficial documents from South Arabia would lead scholars to assume a largely literate society. In fact, inscriptions were not the product of private individuals but of scribal schools, and were thus created in laboratories by professional scribes.

As far as monumental script is concerned, this is proved by the existence of different "styles", which refer not only to script, but also to textual models, and that were the product of the different South Arabian writing schools.

For the texts in minuscule writing and in particular for the letters, it is likely that not only did the sender dictate the text to the scribe (and this is demonstrated by the use of the third person referring to the sender), but also that a scribe read the letter to the recipient.

It can be argued, therefore, that in South Arabia there were two levels of literacy: the first, which is the ability to carve spontaneous graffiti, that was however devoid of any need for communication, and the second, which concerns the existence of a level of written communication for practical or non-practical needs and was typical of scribal schools and a class of literate people, part of both the temple and the political authorities.

Macdonald 2005
Stein 2010