The last 30 years or so has seen the discovery of exciting (the word sensational may not be inappropriate) new material from Yemen: letters and documents that shed new light on everyday life in pre-Islamic Yemen. They add immensely to our understanding of Sabaic, and ancient South Arabia's other languages. The texts challenge the classification of Semitic languages and the rhymed religious hymns anchor the Qur’an more firmly in its Arabian environment. At the same time, these new documents provide important new insight into the history and the origin of the western alphabet.

The first large scholarly publication of material (by Peter Stein) was reviewed in the 2012 volume of this Journal (pp. 466–468) and now can be added another equally ground-breaking study, the volume reviewed here.

Until the 1970s, the only form of writing known from pre-Islamic Yemen were its majestic monumental inscriptions (“musnad”), no doubt the most beautiful letters produced in the Semitic world, and in writing in general, until they were equalled by the similar aesthetics (due to the setting of the letters in a geometrical frame) of the Roman capitalis monumentalis. Today, well over 10,000 of these inscriptions have been published; more are being discovered on a regular basis in both Yemen and Saudi Arabia. The corpus is by far larger than what we have from Ugarit, Phoenicia, in Punic, in old Aramaic, or indeed in (old) Hebrew. Scholars have unfortunately not yet recognised their importance for ancient Near Eastern studies.

To these must now be added an equally impressive number of cursive texts. The largest group (approximately 4,000) is housed in the National Museum in Sanaa; about 2,000 are in the Military Museum in Sanaa, about 800 in the Bavarian State Library in Munich, and a further 300 in Leiden. The University Museum in Sanaa and collections in Riad also have holdings.

Almost all these sticks were found in Nashshan, in the Yemeni Jauf, an ancient royal city the ruins of which are today called al-Sauda’.

After almost 30 years of scholarly efforts, their dating and the development of their writing styles can now be considered as quite securely established. The oldest documents are to be dated around the turn of the 1^st millennium bc (one of the Leiden sticks has been carbon-dated in Oxford, yielding a 95.4% calibrated date of between 1073 and 902 bc). For about 1,600 years, this minuscule writing was the everyday form of written communication in ancient Yemen; its practice continued until the eve of Islam when it was superseded by Arabic and parchment. This writing was called zabur (long u), a word known in pre-Islamic poetry, in the classical Arabic dictionaries (“a Yemeni term”), and in the Qur’an (4,163; 17, 55 and 21,105) where it should of course not be translated as “Psalter”, but simply as “writing”. The texts were mostly written (“satara” or “zabara”) by professional scribes, but signed by the sender.

Zabur was written on wooden sticks, mostly palm-leaf ribs, but also small decorticated branches from other woods, an everyday material available everywhere, anytime, and at no cost. The sticks are about thumb-thick, and vary in length between 10 and 50 cm. Writing is from right to left, in neat lines, incised with a pointed stylus. This form of writing is as smooth as writing with ink on parchment or papyrus, and thus perfectly adapted to alphabetical writing.

A chronology was developed by Jacques Ryckmans based on palaeographical considerations. Maraqten (p. 48 ss.), while refining it, retains its basic structure (archaic, 10^th to 8^th c. bc; classical; early zabur; standard zabur; and finally, late zabur, 4^th to 6^th c. ad), setting it into parallel with the development of the monumental (musnad) script. I personally find Stein's (in ‘Palaeography of the Ancient South Arabian script. New evidence for an absolute chronology’ in Arabian Archaeology and Epigraphy, vol. 24, issue 2 (2013) pp. 186–195) classification more convincing: Stein stresses that the chronology of musnad should be kept separate from the development of the minuscule script. He then proposes four major phases (with subgroups, of course) for the minuscule writing. He notes that there is some overlapping between the phases thus identified, with archaisms continuing into later periods, possibly due to regional and capital/provincial differences. I should underline that there are no differences in absolute dating, between those two scholarly systems.

In the earliest phase, from the emergence of the script in the late second millennium, the letters resemble those of the archaic period musnad; they are unconnected. Stein dates this period from the 11^th/10^th c. to roughly the year 500. The second phase (between ca. 800 and 400 bc) shows a beginning of inclination and curved lines. The third phase (500 to 100 bc) sees the emergence of true cursive letters unrecognisable from their musnad origins. Phase IV is the fully developed cursive. It begins around 200 bc.

Maraqten's book is based on the collection of the Sanaa National Museum. Its 101 texts are a selection dated by the author from the 10^th c. bc to the 4^th c. ad; the last two centuries of zabur writing are not represented in the collection of the Museum. According to the author, these 101 sticks are a representative selection from the material: private letters, economic affairs and business transactions, legal documents, religious texts, lists of names, lists of eponyms, regulations of taxes, and finally school exercises, including alphabets, and a stylus.

The texts are published in the same exemplary way as those from the Peter Stein volumes, each-one with transliteration, translation, dating and grammatical notes. The non-specialist interested in pre-Islamic Arabia will welcome Maraqten's setting of the texts into their wider social or economic environment, and the parallels found elsewhere. For each text, a drawing of the writing is provided, and a photograph, some are in colour. The book begins with a comprehensive introduction to the material. Its 127 pages are a monograph in its own right, with chapters on the language and the formulary of ancient South Arabian letters, on chronology, on social and economic history as it becomes tangible through these documents, and on the role of women in this society: women had full legal capacity and appear quite often as actors in business matters.

We shall now provide a few examples that show the breadth of the subjects covered by the documents:

1. No. 2 is a letter addressed to a mother by her children: your children are fine, please write!, 3^rd/4^th c.

2. No. 4: letter concerning a divorce with the woman claiming custody for her children, 2^nd/3^rd c.

3. No. 5: letter to a woman: yes, the king will find a good appointment for your son! 2^nd/3^rd c.

4. No. 32: receipt for 50 containers of honey, 10^th/8^th c.

5. No. 37: a woman acquitting the fees for her lease, through payment of one “balat” coin, i.e. “Pallas”, the South Arabian owl, 2^nd/3^rd c.

6. No. 63: list of tithes for the god ‘Athtar, 7^th/6^th c.

7. No. 80: a school exercise in calculating, 2^nd c. bc to 1^st c. ad.

The zabur texts, emerging, as they do, around or rather before the turn of the 1^st millennium bc, are an important element in the discussion on the origin of the South Arabian civilisation. They definitely establish its date around the year 1000 bc. As to the question of what caused it, two opinions currently characterise the debate: one (the Italian school) sees the emergence of the majesty that is to become Saba as a gradual development from its Bronze Age antecedents. A majority of scholars do however associate the sudden explosion of new civilisational elements (monumental stone construction in Marib, ashlars, tripartite buildings, planned settlements, shared divinities such as ‘Athtar and Il, i.e. Il-maqahu) to an immigration (of elites) from Greater Syria. Writing is the most important element among these various factors. A few years ago, Sedov has already assigned some words or single letters found on shards excavated in Raybun to the last quarter of the second millennium. The zabur texts with their by now established dating around the turn of the millennium and their obvious connection with the early alphabets of the Levant seem to me the final confirmation for the immigration theory, and its approximate dating towards the end of the second millennium.

As I said above, the new material also enriches our documentation for the South Arabian alphabet. Within the general history of the alphabet, the South Arabian alphabet is of major significance. Unfortunately, it is often not – or only marginally – on the horizon of scholars dealing with this much debated question. I will not go here into a detailed discussion of the origin of the South Arabian alphabet (for this see my recent article in the 2015 issue of the British-Yemeni Society Journal), but a few words about the question are in order.

Almost all the world's alphabets go back to a prototype from the Levant that begins with the letters a, b, g, d, and which is therefore called “abjadiya” in Arabic linguistics. It originated somewhere in Syria, possibly in the 15^th c. bc (for a discussion of the traditional opinion that the development of the alphabet was sparked by an Egyptian-Semitic cultural encounter to be localised in the turquoise-mining community of Sarabit al-Khadim on the Sinai, and the various dating hypothesises proposed for such an event, see my above mentioned article). All the world's alphabets are derived from this prototype, the Phoenician, the Greek, the Roman, ours, the Cyrillic, the Arabic, the Hebrew, the Pahlavi, the Tibetan and the various South Asian (Indian) alphabets. This is due to the fact that an alphabet does not live through its written form, but through its oral memorisation in a school context. It is thus that the abjadiya remained (largely) unchanged when it moved from Semitic to totally different languages and cultures.

But the abjadiya is not the only form of the alphabet! From the very beginning, there was another letter order, which we call “halhamiya” (Maraqten, p. 41s), beginning with the letters h-l-h-m (the first h is a normal h, the second h is the h in “Muhammad”). This is the South Arabian alphabet (and consequently of Ge‘ez and modern Ethiopian). It is very important to note that two other (unrelated) halhamiya alphabets were found in excavations in Palestine and in Ugarit; they are dated to the 14^th and/or 13^th c. bc.

The key for understanding the invention of the alphabet in its two letter orders lies in the following: while there are two letter orders in all the world's alphabets, the forms of the letters in both the abjadiya and the halhamiya are identical. This means that the genius of the Semitic alphabet, its acrophonic value (a from alpu, ox; b from bet, house, etc.) was discovered only once. Because of the above observation that a letter order does not change when an alphabet moves into another language, none of the two letter orders can have branched off from the other. It must therefore have been two persons, two schoolmasters, who invented the letters in a joint effort, but parted ways when they arranged them in a particular order which from that moment onwards perpetuated themselves in the world's two alphabetic orders. The pioneering – others called it “brilliant” – discovery of these connections is due to the Russian scholar, Lundin/Loundine (his latest article was in Mare Erythraeum I, 1997, pp. 9–18). The origin of the South Arabian alphabet can thus be firmly assigned to somewhere in Greater Syria, possibly around the year 1500 bc.

The South Arabian alphabet presents another important characteristic: with its 29 letters, it is the only Semitic alphabet that has retained the signs for all 29 proto-Semitic phonemes (Ugaritic has 27 letters, Phoenician and Hebrew 22). This supports a rather early dating for the South Arabian alphabet, and for its (not documented) abjadiya counterpart. The invention would have occurred in an environment (somewhere in Greater Syria) where (proto-) Aramaeans and proto-Sabaeans were still somehow united or in close contact: The verbal stem system, the most important morphological marker in Semitic languages, is the same in Aramaic and Sabaic, and distinguishes them from all other Semitic languages. Old Aramaic also had the full inventory of the 29 proto-Semitic phonemes (but not yet an alphabet), see Kottsieper and Stein, in ‘Sabaic and Aramaic – a common origin’ in Supplement to the Proceedings of the Seminar for Arabian Studies 44 (2014), pp. 81–88.