Matara, near Senafe, Eritrea
The Hawulti of Matara
Stele placed in the 3rd or early 4th century (a disc and crescent, and a South Arabian or local hand)
At Matara, near Senafe, a stele about 5 metres high carries a disc set on a crescent and four lines of writing at eye level. The letters have been dated to the 3rd century, and also to the early 4th, before the cross replaces that heavenly mark. The same town has older South Arabian inscriptions built into later walls. One reading calls this a colony. Another calls it a local centre that took up a foreign script and a foreign sign.
Why this is extraordinary
The stele at Matara stood in two pieces by the hill of Goual Saim until the 1920s, when it was mended and set on the track toward Berakit. It is pantellerite, a pale volcanic rock, about 5 metres from the ground in the excavation account and 5.5 metres in another. Near the top someone cut a disc sitting on a crescent. At eye level, four lines say that AGZ raised the stone for his ancestors, and that friends, named in the text, carried it. The photograph is that reset stone in the plain. The device is on the upper face. This frame does not give you a clean circle and a clean crescent, so the name of the mark comes from the publications, not from a trick of the light in the picture.
What the mark means is published twice, in the same calm tone. One account calls it the symbol of Astarte, a device already used on stelae, incense burners, and coins from the early pre-Aksumite centuries through the 3rd century, and replaced by the Christian cross in the 4th. The letter shapes, in that account, date the monument to the 3rd century. Another account gives the disc and crescent to Šams and Sin, a sun goddess and a moon god, and says the mark was there to put the stone under their protection. The letters, still without vowels, put that reading in the early 4th century. Both are sky readings. They name heavenly bodies as a sign cut in stone. They do not claim that the stele’s long face points at moonrise or at a solstice. No azimuth is published. You learn the sign by recognising the disc and the crescent, not by waiting for a dawn to strike the stone.
The town is larger than the stele. The upper layer, when it was dug, held four residences, two churches, and about thirty ordinary houses, dated by pottery and coins to the 6th and 7th centuries, with two buildings that may be 4th or 5th. In those walls were reused stones engraved in South Arabian script, some of them perhaps as old as the 8th century BC. For a long time that kind of script was taught as proof that kingdoms across the Red Sea had colonised the highlands. A closer look at the texts found a local society already in place, taking on a foreign writing and a foreign sign. What the regional argument does not find is a South Arabian king named in the pre-Aksumite inscriptions, or one uninterrupted South Arabian province. About 100 kilometres away, the Ona settlements near Asmara kept their own things beside centres that look more foreign. A 15th-century manuscript calls these ruins the place King Kaleb destroyed, coming from Aksum by a tunnel. That is a story about a link to the capital. It is centuries later than AGZ.
The catch is the resetting. The stone we see was stood up in the 1920s. Its original socket is not the hole a visitor might imagine under it. What this changes: if the disc and crescent are a local patron’s appeal to sun and moon, or to Astarte, Matara’s stele is highland ancestor-work in a shared Red Sea language. If the whole town is a colony, the same stone is an outpost’s god. The port of Adulis is the coast below this plateau, and it is a different room. No measurement here turns the stele into a sighting instrument.
Why it matters
If the stele is a South Arabian monument on African ground, the highland towns begin as outposts. If local rulers cut it for their own dead, the disc and crescent are a sign those rulers chose, and the script is a tool, not a conquest.
The stone is pantellerite, a trachyte. It lay in two pieces by the hill called Goual Saim until the 1920s, when it was mended and set up beside the track to Berakit. From the ground it is about 5 metres; another account says 5.5. At the top is a disc on a crescent. One publication names that device as the symbol of Astarte and notes that the same mark appears on stelae, incense burners, and coins from the early pre-Aksumite centuries until the 3rd century, and that the cross replaces it in the 4th. Another reading gives the disc and crescent to Šams, a sun goddess, and Sin, a moon god, as protection over the stone, and dates the letters to the early 4th century because they are not yet vocalised. The four lines say the stele was raised by AGZ for his ancestors and carried by friends whose names are listed. Palaeography of the letter shapes is what puts it in the 3rd century in the excavation report. The town around it is older and younger than the stele. Excavation of the upper layer found four residences, two churches, and about thirty ordinary houses, the pottery and coins of the 6th and 7th centuries, with two buildings possibly of the 4th or 5th. Stones with South Arabian inscriptions, some of them monumental and some perhaps of the 8th century BC, were reused in those later walls. A local name for the ruins, the ruins of King Kaleb, comes from a 15th-century manuscript in which the king comes from Aksum and destroys the town. That is a tradition about a link to Aksum. It is not the date of the stele. No one has published an azimuth of the stone’s long face. The disc and crescent are heavenly bodies cut as a sign, not a sightline to moonrise.
Photo tour
1 photograph, unedited. On a phone, swipe the row. Choose one to look closely.
,_stele_axumita_05.JPG){kind=link}
- The photograph does not make the disc and crescent crisp. The publications are the source for that identification.
What is still unexplained
These sit beside the consensus account, with the same visual weight. None of them is marked solved.
Location and dating
- Location
- Matara, 1.5 kilometres into the plain below the hill Goual Saim, near Senafe, Debub, Eritrea. The pin is 14.688 north, 39.419 east. The stele was moved from where it lay in two pieces and set up by the track in the 1920s. The photograph is that resetting.
- Period label
- Current consensus — open to revision The stele is placed in the 3rd century by one palaeography and in the early 4th by another. The upper town is 6th and 7th century, with hints of the 4th or 5th, and reused inscriptions that may reach the 8th century BC. The disc and crescent give way to the cross in the 4th century on the region’s monuments.
- How this dating was done
- Palaeography of the four lines, not radiocarbon. The town’s upper layer is dated by pottery, coins, and small inscriptions. The South Arabian pieces are dated by their own script and then found out of place, reused.
- What could change it
- A radiocarbon sample sealed against the original setting of the stele would replace the letter date. The original setting was disturbed when the stone was stood up again.
What is actually there
- The stele is about 5 metres, or 5.5 in one account, of pantellerite.
- A disc on a crescent is carved at the top. Four lines name AGZ, his ancestors, and the friends who carried the stone.
- One reading calls the device Astarte’s and dates the letters to the 3rd century. Another calls it Šams and Sin and dates the letters to the early 4th.
- The cross replaces that device on regional monuments in the 4th century.
- South Arabian inscriptions were reused in the later walls.
- No azimuth of the stele is published.
Open questions
| Question | Status |
|---|---|
| Which gods does the disc and crescent name? | Open. Astarte, and the pair Šams and Sin, are both published. Both are sky readings of a sign, not of a horizon. |
| Did people from South Arabia found Matara? | Open. The script and the reused stones are real. A named South Arabian king, and a continuous colony, are not what the highland inscriptions show. |
The current narrative
Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.
The current narrative — open to revision
A local stele using a shared heavenly sign
AGZ raised the stone for his ancestors in the 3rd or early 4th century. The disc and crescent are a Red Sea sign for heavenly protection, cut by or for a highland patron, not a proof that a South Arabian kingdom owned the town.
What supports this
- The text names AGZ, ancestors, and the carriers, not a foreign king.
- The same device is already on local coins and incense burners across centuries.
- Regional epigraphy finds influence on a population that was already organised, and Ona sites nearby stay local.
What pushes back
- The letter forms and the device are South Arabian in origin.
- Older monumental South Arabian inscriptions were built into Matara’s later walls, so people who used that script were here before the stele’s century.
- A 15th-century manuscript still explains the ruins by a king coming from Aksum, which shows how readily the town is given an outside founder.
What would settle it
It is the centre because the inscription’s own grammar is a local dedication, and the colony reading has been narrowed by the absence of a South Arabian king in these texts. It moves if AGZ is shown to be a title from across the sea rather than a name at Matara.
Emerging narratives
Other published readings. Each may be right. None is adopted here unless the evidence carries it.
An emerging narrative
A South Arabian colony’s monument
The stele and the older inscriptions are the trace of South Arabian settlement, and the disc and crescent are those settlers’ gods.
What supports this
- Script, divine sign, and reused 8th-century-style stones are a real South Arabian package.
- The older teaching of the highlands was dominance by Maʿin and Saba.
- Disc on crescent is at home in South Arabian religion, whether as Astarte or as sun and moon.
What pushes back
- No South Arabian king is named on the pre-Aksumite inscriptions of the highlands in the survey used here.
- The upper city is Aksumite, centuries later, with churches. A colony would need a continuous town, and the excavation of the upper layer does not supply one back to the 8th century BC.
- The stele text is about ancestors and friends, which reads as a family monument.
What would settle it
It stays because the script did come from South Arabia and the stones prove contact. It is not the centre while the dedication is local and the regional texts do not name a colonising king.
Should the current view be reevaluated?
Open to revision
The stele is a real inscribed stone with a heavenly sign and a letter date in the 3rd or early 4th century. Who commissioned that sign, and which gods it names, are published in more than one way. Open is the state of the reading, not of the stone’s existence.
What drives this
- A disc and crescent read as Astarte, and also as Šams and Sin.
- A 3rd-century palaeography and an early 4th-century one.
- South Arabian script in a town whose founders are argued as colonists and as local patrons.
This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.
Weigh it yourself
Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.
Visitor lean so far — not evidence. On this browser only; a shared tally is not collected.
If the contested reading is right
What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.
A local ancestor stone under a shared sun-and-moon sign
Matara’s stele would be highland work in a Red Sea symbolic language, and the South Arabian stones in the walls would be contact and reuse rather than a colony’s foundation deed.
Why it is still worth a test
A reading of AGZ that is agreed, and a record of the stele’s original socket, which the 1920s resetting removed.
Sources
- Matara, the archaeological investigation of a city of ancient Eritrea Palethnologie, the stele, Astarte, and the 3rd-century letters
- Hawulti The sun-and-moon reading and the early 4th century
- Colonies and colonialism in the Horn of Africa Azania, the colony argument and its limits