Marginal Notes on: BS 164 – the tribe of ʿād.

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions.

The inscription BS 164 was discovered by Michael Macdonald and Ali Al-Manaser on the “OCIANA Badia Survey” of 2015 and published on OCIANA as follows:

Photograph of BS 164 (courtesy OCIANA).

l ms¹k bn zmhr bn yzr bn tʾm w wgm ʿl- ʾb -h w ʿl- ʾḫ -h {ḏ} ʾlʿd

By Ms¹k son of Zmhr son of Yzr son of Tʾm and he grieved for his father and for his brother who ʾlʿd

The edition took the final four letters as a verb, but did not supply a translation. I could not find any root lʿd in Arabic or Northwest Semitic. I believe, however, that these final letters do not constitute a verb, but rather the words ʾl /ʾāl/ ‘lineage’, ‘people’ and the group name ʿd, which can be nothing other than ʿād, the ancient Arabian tribe mentioned in the Qur’an. I would suggest the following reading and translation:

l ms¹k bn zmhr bn yzr bn tʾm w wgm ʿl- ʾb -h w ʿl- ʾḫ -h {ḏ} ʾl ʿd

By Ms¹k son of Zmhr son of Yzr son of Tʾm and he grieved for his father and for his brother of the lineage of ʿd

This is the first attestation of this group name in Safaitic, but the tribe is known already in Hismaic. For example, TIJ 004 (King 1990: 650–651) contains a signature of a man from the lineage of ʿd:

l zhy bn ʿmr ḏ- ʾl ʿd

By Zhy son of ʿmr of the lineage of ʿd

Another Hismaic inscription published by Farès and Zayadine (1998) mentions the building of a temple for the Arabian goddess Allāt by the tribe of ʿād.

These attestations of ʿād in northwest Arabia and the southern Levant contradict the narrative of medieval Islamic historians, who put ʿād in southern Arabia, and underscores the unreliability of the “origin stories” found in such works. The attestation of ʿād in Northwest Arabia also supports its association with ʾiram, which is modern Wadi Rum. The toponym is attested as ʾrm /ʾiram/ in Nabataean as well (e.g. Savignac 1933).


Farès, S. and Zayadine, F. (1998) “Two North-Arabian inscriptions from the temple of Lât at Wady Iram”, Annual of the Department of Antiquities of Jordan (ADAJ) 42: 255-258.

King, G.M.H. Early North Arabian Thamudic E. A preliminary description based on a new corpus of inscriptions from the Ḥismā desert of southern Jordan and published material. Ph.D thesis, School of Oriental and African Studies, University of London, 1990. [Unpublished]. 1990. Pages: 650–651

Savignac, M. Raphael 1933. Le sanctuaire d’Allat à Iram (1). Revue Biblique, 42: 405-422, pl. 24.

[TIJ] Harding, G.L. & Littmann, E. Some Thamudic Inscriptions from the Hashemite Kingdom of Jordan. Leiden: Brill, 1952. Pages: 9 Plates: II

Marginal Notes on: HASI 9-11

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions.

The inscriptions HASI 9-11 are carved on a single stone, discovered by Abdul-Qader Al-Husan in the Mafraq Province, Jordan and published on OCIANA. The rock is said to be in situ but no further details about its location are given. The inscriptions seem to have been carved by a single hand. OCIANA reads the texts as follows:

HASI 9-11 (courtesy OCIANA).


HASI 9: l ns²ḥ bn hknf bn ydʿ

HASI 10: l ʾḥḥt bn ns²ḥ bn hknf

HASI 11: l ṣḥr bn hknf bn ydʿ w rmy gddʿ

The men Ns²ḥ and Ṣḥr are brothers and ʾḥḥt is the son of Ns²ḥ. While it is common to regard these texts as signatures, the fact that all three texts seem to be carved by the same person permits other interpretations as to their purpose. But before getting to that issue, I will discuss the interpretation of HASI 11. OCIANA has translated the text as:

By Ṣḥr son of Hknf son of Ydʿ and he went to a level plain

While the commentary does not discuss how this meaning was achieved, one may assume that the editor interpreted rmy as Classical Arabic ramā ‘to repair to a place’ (Hava, 264b) and gdd as Classical Arabic ǧadad ‘level ground’ (Lane, 386c). The final ʿ seems to be omitted from the interpretation. The letter is not on the photograph, and so the reading of the final word should be amended to gdd.

While it is true that the Safaitic words correspond to the Arabic ones identified and the result produces a grammatical sentence, the reason to carve such a pointless statement into rock is hard to find, even more so given that the author does not specify a specific plain but uses the indefinite form. The Safaitic inscriptions abound with what appear to be expressions of profane activities, such as pasturing and migrating, but these nonetheless would have held much significance to the inhabitants of the Ḥarrah and are essential to their survival. A prayer for security while pasturing therefore is far from pointless. This interpretation on the other hand seems only to state a matter of fact and, while possible, I wonder if the text could be better interpreted in another way.

The Safaitic inscriptions are highly formulaic in their contents but can vary significantly lexically. Inscriptions requesting security while keeping watch are common, but the verbs for ‘guarding’ or ‘keeping watch’ can vary from ḫrṣ, rʾy, ḫll, nẓr, tẓr, and so on; some of these are rather common while others are rare.

A common phrase, attested 46 times according to OCIANA’s count, is tẓr mny. This has been translated in various ways by different editors. Winnett and Harding give the translation ‘he awaited Fate’, referring presumably to death, and this is the only translation found in OCIANA in its present form. This translation contradicts the depiction of deified Fate as an unseen force that snatches its victim at any moment – not quite something one would or could await. I suggested in 2015 that the subject of tẓr was in fact mny ‘Fate’ and the phrase should better be translated as ‘and Fate lay in wait’. This better explains the prayers for deliverance and salvation that often follow:

RSIS 309: tẓr mny w yṯʿ rwḥ ‘Fate lay in wait so Yṯʿ send ease!

ASWS 169: tẓr mny f h rḍw flṭ m-bʾs¹ ‘Fate lay in wait so, O Rḍw, deliver (him) from misfortune’

WAMS 19.2: tẓr mny f h rḍw flṭ-h ‘Fate lay in wait so, O Rḍw, deliver him’

Now, why is this all relevant? I do not think the present inscription is a paraphrase of tẓr mny, but I suspect that gdd is another term for deified Fate, equivalent perhaps to mny ‘Fate’ and mt ‘Death’.

The term al-ǧadīd can refer to ‘a thing of which one has no knowledge’ and as such is used as an epithet of ‘Death’ (Lane, 387a). The following excerpt from the lemma on this root in the Lisān (s.v.):

الجديد: ما لا عهد لك به, ولذلك وصف الموت بالجديد.
هذلية: قال أبو ذؤيب:
فقلت لقلبي: يا لك الخير! إنما
يدليك للموت الجديد جبابها

The epithet may further derive from the basic sense of the root gdd meaning ‘to cut’ (Classical Arabic ǧadda-hu ‘he cut it off’ (Lane, 384b), a common metaphor for Fate, e.g. one’s “lot” in English or Kismet “fate, destiny”, from Turkish, ultimately from Arabic “portion”, from the root qsm ‘to divide’.

If we take gdd as a term for Fate, then the verb rmy is best understood as ‘to throw, cast, shoot’, with the ellipsis of its object. This verb has previously been attested in Safaitic (Al-Jallad 2015: 339). The motif of deified Death/Fate shooting arrows at the doomed is common in Classical Arabic poetry. To give just an example, consider the following distich by  Nābighah al-Dhubyānī (d. 604 CE):

أني وجدت سهام الموت معرضة بكل حتف من آجال مكتوب
‘I have indeed seen the arrows of Death directed at every doomed man from a pre-destined decree’ (al-Nābigha al-Dhubyānī, wa-ajmal ʾashʿārih, comp. ‘Amr Yūsuf (Alexandria, n.d.), 27).

Returning to the purpose of these texts, it is possible that the Ṣḥr had died and Ns²ḥ and his son buried and grieved for him, carving these texts as a memorial, the group of texts being carved by one of these individuals. It is also possible that all three men died and the texts are memorial, the epitaph applying to all of them.

The new translation I cautiously suggest is therefore:

HASI 11: l ṣḥr bn hknf bn ydʿ w rmy gdd 

‘By Ṣḥr son of Hknf son of Ydʿ and Fate shot (its arrows)’.


All sigla follow the OCIANA identifications.

Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (SSLL 80). Leiden: Brill.

Marginal Notes on: C 4717

Inscription C 4717, known only from a hand copy, reads as follows:

l ḫyḏt bn ḫbṯ bn s¹m(k) bn s¹wr bn mlk w rʿy h-nḫl bql w ngʿ b-ry s¹rqt f h bʿls¹m[n] (ġ)[n](y)t

Tracing Dunand 925 (courtesy of OCIANA).

The translation of most of the text is straight forward:

‘By yḏt bn ḫbṯ bn S¹m(k) bn S¹wr bn mlk and he pastured in the valley on fresh herbage and grieved in pain b-ry s¹rqt  so, O Bʿls¹m[n], let there be [abundance]’

The section in bold occurs only once in the corpus of Safaitic. Ryckmans, the original editor, translated it as “he wept because the land was watered”, but this interpretation is hard to justify logically and linguistically. The first component resembles the well-known formula b-rʾy ‘at the rising’, referring to the rising of asterisms on the horizon (Al-Jallad 2014, 2016). If a glottal stop can be restored here, then it would suggest that the following term refers to an asterism or heavenly body.

The root s¹rq concerns the semantic domain of theft, attested in Arabic. None of the zodiacal or para-zodiacal constellations can be reconciled with this meaning. It is possible that the term is the name of an asterism that does not survive in any known tradition, and if that is the case then it is impossible to identify it.

The Robber is one of the names of Mars in the Mesopotamian star catalogs: mul (see Koch-Westenholz 1995). Is it possible that the Safaitic actually attests a name of Mars? Was its rising considered a bad omen, at least at this particular time? Another name of Mars, The Evil One, mul lu-um-nu, is possibly attested in Safaitic. I suggested (Al-Jallad 2014: 225–226; Al-Jallad 2016: 87, 102) that the phrase ʾlmn b-ʿqbt (KRS 1551) refers to an astronomical event, namely, the passage of Mars through Scorpio and seems to have been considered a bad omen, as the author then prays for relief or rain. Could there have been multiple names for the planet Mars as in the Mesopotamian tradition?

It is possible to venture a non-astronomical interpretation of this text by taking s¹rqt as thieves, cf. CAr saraqatun. The phrase could then be translated as ‘he grieved when seeing thieves’, but such an expression would find no parallel so far in Safaitic, at least with b-ry.

For now, the astronomical interpretation seems to be the best fit for this difficult text. But as with all inscriptions known only from hand copies, a great deal of caution must be observed when advancing theories based on them. Until the rock is rediscovered, we cannot know for sure what this text truly says.



Al-Jallad, A. 2014. An ancient Arabian zodiac. The constellations in the Safaitic inscriptions, Part IArabian Archaeology and Epigraphy 25.2: 214–230.

Al-Jallad, A. 2016. An Ancient Arabian Zodiac: the constellations in the Safaitic inscriptions, Part II. AAE 27: 84–106

[C] Ryckmans, G. Corpus Inscriptionum Semiticarum: Pars Quinta, Inscriptiones Saracenicae Continens: Tomus I, Fasciculus I, Inscriptiones Safaiticae. Paris: E Reipublicae Typographeo, 1950–1951.

Koch-Westenholz, U. 1995. Mesopotamian astrology: an introduction to Babylonian and Assyrian celestial divination. Copenhagen: Museum Tusculanum Press.

[KRS] Inscriptions recorded by Geraldine King on the Basalt Desert Rescue Survey in north-eastern Jordan in 1989 and published here

Marginal Notes on: Inscription 259 in Khraysheh 2002

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions.

Fawwāz Khraysheh published in 2002 an edition of Safaitic inscriptions from Biyār al-Ghusayn, Jordan. The 512 texts contain a good mix of genres, mostly conforming to the established formulae, but attesting a few new lexical items and constructions. A few of these texts have figured into my grammar (2015), and more will be used in the second edition. My reluctance to draw on this particular edition comes from the fact that there are virtually no photographs of the inscriptions in the book. While such a weakness is forgivable for older editions, there is no reason for such an omission in modern works. Thus, we are made to rely on the tracings of the author. Even the most experienced epigraphist can trace glyphs incorrectly, as a large number closely resemble each other even in ideal conditions, and so drawings are not the best source to establish the attestation of new vocabulary or constructions. Khraysheh passed away nearly a decade ago and the location of his photographs is unknown. It is hoped that his estate can make his field material available to scholars in the future or that his texts will be re-discovered by a future expedition to Biyār al-Ghusayn.

The inscription of interest is no. 259, which Khraysheh draws as follows:


Khraysheh’s reading and translation:

KhBG 259

l ʾḫʾl bn gd h-dr w ṣyr b- ḏʿṯr m- m[dbr]

‘this is the campsite of ʾḫʾl son of Gd and he set off in the month of Ḏū ʿAṯṯar from the desert’

In his short commentary he states:

“It is not strange that some of the Safaites would use the names of the South Arabian months, despite the fact that this is the first time that ḏʿṯr is attested as a month name in Safaitic” (my translation)

In fact, it is rather odd that a South Arabian month would appear in the Safaitic inscriptions, given that there is no evidence for South Arabian influence otherwise. Moreover, at least three Babylonian months have been attested so far, knn (=kānūn), s¹bṭ (=šubāṭ) and nṣn (= nīsān). Month names, however, are exceptional and the normal way of reckoning time is through the use of a zodiac star calendar (Al-Jallad 2016).

Assuming that Khraysheh copied the inscription correctly, ḏʿṯr can refer to a month or celestial body connected with the deity ʿAṯtar. ʿAṯtar is a proto-Semitic deity, worshipped in South Arabia as ʿṯtr, but also found in Ugaritic as ʿṯtr, Mesopotamia Ishtar, Canaanite, ʿštr, and Aramaic ʿtr. Thus the presence of this name alone does not require a connection with South Arabia. ʿAṯtar has not yet –to my knowledge– appeared as a deity in the Safaitic inscriptions, so at the very least the term would seem to be foreign.

A month with the divine name ʿṯtr occurs in the Minaic calendar as ḏʿṯrt and in Sabaic as ḏʿṯtr (see Beeston 1956). Neither for these names matches perfectly the Safaitic attestation and therefore there is no a priori reason to assume a borrowing from one of these calendars. While the change of ʿaṯtar to ʿaṯṯar is trivial, there is no reason to assume that it is necessary in Safaitic, nor is there any reason to assume the deletion of the final t if the form was borrowed directly from Minaic. Thus, to account for these oddities we must appeal to one or more ad-hoc phonological changes during the borrowing process, or a pronunciation that differed from the spellings of the Ancient South Arabian texts.

In possible support of the South Arabian origin is the use of the relative pronoun with the month name. This practice extended to calendars outside of South Arabia as well: some of the month names of the Arabo-Islamic calendar contain this pronominal element, e.g. ḏū l-ḥiǧǧah, and possibly the calendar of Ancient Dadān (this will be argued in the Leiden University thesis of Fokelien Kootstra).

Based on the Safaitic evidence, ʿAṯtar was not worshipped in the Ḥarrah and the deity is not mentioned by ibn al-Kalbī in his Book of Idols. As far as I know from the Thamudic inscriptions, none of the variants of ʿAṯtar has appeared in prayers. However, at Dūmat al-Ǧandal the deity ʿtrs¹m /ʿattar-samā/ ‘ʿAttar of the heavens (lit. sky)’ is encountered in a single text:

h rḍw w nhy w ʿtrs¹m sʿd-n ʿl wdd-y

O Rḍw and Nhy and ʿtrs¹m help me in the matter of my love!

ʿtrs¹m is attested in a few Thamudic inscriptions from North Arabia (e.g. Winnett and Reed 1973: 136, 160, 161). Positing a connection with the present attestation, however, is problematic for the same reasons as the South Arabian connection – the name is not an exact match.

Considering all of the evidence, I believe that Khraysheh’s suggestion of a South Arabian origin to be the most plausible, but if so, the source was most likely the Minaean trading colony at Dadān in North Arabia rather than South Arabia proper. This month name could have diffused to surrounding Arabian tribes, like the Ḥawīlat, Liḥyān, or Ruhay. During the diffusion process, the final t could have been lost – perhaps because it was subject to common sound change of at to ah and ultimately ā. While this did not happen in Minaic or Safaitic, it did happen in the local Arabian dialect of Dadān, at least in some conditions. Just as the Babylonian month names diffused to the Safaitic writers of the Ḥarrah, nomads of North Arabia may have taken over the names of months from oases such as Dadān and Taymāʾ.

While the North Arabian tribes I have mentioned above do not usually carve Safaitic inscriptions, contact with them is well attested in the Safaitic corpus, e.g. WH 736.a (Al-Jallad 2015:60). Thus, the Safaitic writer could have learned this month name from interaction with North Arabian tribes as they migrated into the Ḥarrah, a phenomenon also well attested in the Safaitic inscriptions. It is also possible that the author belongs to one of the North Arabian tribes I have mentioned above – Safaitic inscriptions were composed by diverse groups and a few inscriptions were in fact composed by men from the Ḥawīlat tribe of North Arabia (Al-Jallad 2015: 18-21).

I would suggest the following translation for the text:

l ʾḫʾl bn gd h-dr w ṣyr b- ḏʿṯr m- m[dbr]

‘By ʾḫʾl son of Gd, in this region, and he returned from the inner desert to a place of water during (the month of?) Ḏū ʿAṯṯar’.


Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (SSLL 80). Leiden: Brill.

Al-Jallad, A. 2016. An Ancient Arabian Zodiac: the constellations in the Safaitic inscriptions, Part II. AAE 27: 84–106

Beeston, A.F.L. 1956. Epigraphic South Arabian calendars and dating. London: Luzuc.

Khraysheh, F. 2002. Nuqūš ṣafawiyyah min biyār al-ġuṣayn. Irbid: Yarmouk University Press.

Winnett, F.V. and W.L. Reed. 1973. An Archaeological-Epigraphical Survey of the Ḥāʾil Area of Northern Saʿudi Arabia, 22: 53‒114.


Marginal Notes on: Safaitic inscriptions published in The Semitic Languages: An International Handbook

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions.

In the 2011 The Semitic Languages: An International Handbook, Hayajneh published in his contribution on Ancient North Arabian two new Safaitic inscriptions (p. 775, fig 44.9a). The first is in a mixed Safaito-Hismaic script with square letter-shapes. This hand is typical of the lineage of ʿmrt. The texts come from northern Jordan, but no further details of the location are given.

Hayajneh (2011: 775, fig 44.9a; photo H. Hayajneh)

Hayajneh reads and translates the inscriptions as follows:

” l tm bn ṣrmt ḏ l mrt w rʿy ḍrk w ʾlt w dšr ġnyt

‘By Tm son of Ṣrmt of the tribe ʿmrt and he pastured Ḍrk (place name!). Oh Lt and Dšr (grant) wealth’. ”

The reading is sound, but I would restore an alif between the and l. Its absence is most likely the result of a writing error, as the loss of the glottal stop would yield a glide in this position. In fact both ḏwl /ḏū-āl/ and ḏyl /ḏī-āl/ are rarely attested. I would, however, like to suggest a better translation for ḍrk. Unattested toponyms are generally a last resort for the interpretation of these texts, since anything can be a toponym and context rarely rules out such an interpretation. Considering that prayers are often connected to the narrative component, I would instead suggest that ḍrk should be interpreted along the lines of Classical Arabic ḍaruka ‘to be struck by misfortune’; ḍarīk ‘poor, hungry’. In the context of pasture, I suggest that the word refers to scarcity, perhaps on account of a drought. Syntactically, it is an accusative of circumstance, so the phrase should be translated as /raʿaya ḍarīka/ ‘he pastured suffering from scarcity’. The prayer then petitions the Nabataean deity ds²r Dusares and the goddess ʾlt (Allāt) to alleviate his condition, asking for ġnyt /ġaniyyat/ ‘abundance’. The new translation is thus:

New translation

By Taim son of Ṣrmt of the linage of ʿmrt and he pastured suffering from scarcity so, O Dusares and ʾAllāt, let there be abundance!

The second inscription consist only of a name, which the edition reads as l ʾs¹lm bn ṣmʿn. Note that the strongly resembles an r, but the personal name rmʿn is not known.



Hayajneh, Hani. 2011. “Ancient North Arabian,” in Stephan Weninger with Geoffrey Khan, Michael Streck, and Janet Watson (eds.), The Semitic Languages: An International Handbook (Handbücher zur Sprach- und Kommunikationswis-senschaft 36). Boston-Berlin. 756–782.

Marignal Notes on: “A POSSIBLE ATTESTATION OF THE NABATAEAN MINISTER SYLLAEUS IN A NEW ANCIENT NORTH ARABIAN (SAFAITIC) INSCRIPTION” by Nada Al-Rawabdeh and Sabri Abbadi, in Mediterranean Archaeology and Archaeometry, Vol. 16 No 2, (2016), pp. 33-40

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions.

Rawabdeh and Abbadi publish the second Safaitic inscription that perhaps mentions the famous Nabataean minister Syllaeus. The inscription is basically identical to one already published by Abbadi in 2001 (see references). The article contains a balanced philological discussion of the new text, and concludes with some remarks on the historical event to which the two Safaitic inscriptions could refer. I will not venture into historical speculation here, as the references in these texts are too brief to say anything for certain, but will rather offer some improvements to the interpretation of the two inscriptions, and notes on the translations of previously published inscriptions in their edition.

The S¹ly inscription published in 2001 by Abbadi is republished in this edition, with the following reading and translation:

Photo from Abbadi 2001
Photo from Abbadi 2001

Original Reading and Translation

l tm bn ms¹k bn qtl bn brd bn ḥmt w wgm ʿl- ġyr w ʿl- qtl w ʿl- mṭl s¹nt ʾty s¹ly m- rm w ḫrṣ h-s¹nt f h bʿls¹mn ġwṯ w s¹lm w qbll l- ḏ ʾḥb

“By Tm son of Ms¹k son of Qtl son of Brd son of Ḥmt and he grieved for Ġyr and for Qtl and for Mṭl the year S¹ly came from Rm and he kept watch this year and so O Bʿls¹mn [grant] help and security and [show] benevolence for whoever ʾḥb.”


Revised Translation

‘By Tm son of Ms¹k son of Qtl and he grieved for Ġyr and for Qtl and for Mṭl, the year S¹ly came from Rome and he kept watch this year, so O Bʿls¹mn, remove affliction so that he may be secure; and may there be a reunion with him/those whom he loves’.

Notes: The edition follows the old translation of qbll as ‘benevolence’, whereas it should be understood as ‘reunion’, ‘reunification’ (Al-Jallad 2015: 333) given its textual context. This noun or infinitive occurs most frequently following expressions of longing or keeping watch for loved ones who are away, e.g. ts²wq ʾl- ‘he longed for’ and ḫrṣ ‘kept watch for’, and so the logical prayer in such cases would be for reunification with absent loved ones. Morphologically, it is an R -stem, cognate in form with Classical Arabic iqballa, perhaps vocalized as */eqbelāl/ or */qeblāl/.

The term ʾḥb is previously attested as an elative in the expression ts²wq ʾl- h-ʾḥb ‘he longed for the most beloved’…[f h lt] qbll ‘so O Lt, may there be a reunion’. The occurrence of ʾḥb with qbll then is a precedent for the present inscription, and motivates us to understand the occurrence of ʾḥb here as an elative or perhaps as a suffix-conjugated verb /ʾaḥabba/ ‘he loved’. This verb is attested as ʾḥbb ‘he loved’ elsewhere (Al-Jallad 2015:321). The relative pronoun may have a singular or plural referent.


The new S¹ly inscription (2016)

Photo and tracing from Rawabdeh and Abbadi 2016
Photo and tracing from Rawabdeh and Abbadi 2016

Original Reading and Translation

l tm bn ms¹k bn qtl bn brd bn ḥmt bn ġlmt bn mr bn ʾfty bn gml w wgm ʿl- ġyr w ʿ l- mlṭ w ʿ l- qtl s¹nt ngy s¹l[y] mn rm w ḫr{ṣ } h- s¹nt f h bʿ ls¹mn ġw{ṯ } {w}{s¹}{l}{m} {w} {q}{b}{l}{l} {l}- {ḏ } {ʾ }{ḥ }{b}

“By Tm son of Ms¹k son of Qtl son of Brd son of Ḥmt son of Ġlmt son of Mr son of ʾfty son of Gml and he grieved for Ġ yr and for Mṭl and for Qtl the year [S¹ly] fled from Rm and {he kept watch} this year and so O Bʿls¹mn [grant] {help} {and} {security}{and} [show] {benevolence} {for} {whoever} {ʾḥb}”.


Revised Translation

“By Tm son of Ms¹k son of Qtl and he grieved for Ġyr and for Qtl and for Mṭl, the year [S¹ly] fled from Rome and he kept watch this year, so O Bʿls¹mn, {remove affliction so that he may be secure; and may there be a reunion with him/those whom he loves}”.

Notes: See the discussion of the previous text, which is basically identical to this one, for notes on the revisions. As the ed. pro. points out, the equation of ngy in this text with ʾty ‘he came’ of the previous one, suggests it carries the meaning ‘he fled, escaped’ rather than ‘he was announced’. Both meanings are attested (Al-Jallad 2015: 331).

It is remarkable that the same elaborate inscription was produced nearly identically twice. This certainly gives us something to think about when it comes to the production of these texts.


The article also cites several previously published Safaitic inscriptions with erroneous translations (see the edition for references). I will rectify these here (corrections in bold):

C 742: l s¹r bn nẓr bn ṣhyn bn gʿl bn rs¹l w {n}fr m- rm s¹nt ws¹q ḏ- ʾl rhy nbṭ mġwt f h lt s¹lm w [n]qʾt l- ḏ {y}ʿr

Revised Translation: “By S¹r son of Nẓr son of Ṣhyn son of Gʿl son of Rs¹l and he escaped from the Romans in the year that those of the lineage of Rhy clashed with the Nabaṭaeans at mġwt, so, O Lt, may he be secure; and may whosoever would efface (this writing) be thrown out (of the grave)”.

mġwt = this word is unattested in Safaitic and may be a copyist error, as we do not have the photograph of this inscription. The best suggestion at the current moment is to take it as a toponym.

WH 2815: l ʿbd bn {y}ġṯ ḏ- {ʾ}l {b}{s¹}ʾ w ngy m nf{r}t w ʾḫ -h s¹nt mrdt nbṭ ʿl- ʾl {r}m f ʾt s¹lm

Revised Translation: (missing in Rawabdeh and Abbadi 2016): “By ʿbd son of Yġṯ of the lineage of ʾl Bs¹ʾ and he and his brother were announced (leaders) over a company of men, the year the Nabataeans rebelled against the Romans, so may peace come.”

ANSWS 79: l ẓʿn bn grmʾl bn ẓʿn bn bnt bn ẓʿn bn ḫṭst ḏ- ʾl kn w wgd s¹fr grmʾl f bʾs¹ mn ẓll w qnṭ ʾl rm s¹nt yhd f h lt….. wqyt m bʾs¹

Revised Translation: “By Ẓʿn son of Grmʾl son of Ẓʿn son of Bnt son of Ẓʿn son of Ḫṭs¹t of the lineage of Kn and he found the inscription of Grmʾl, for those who remain despair; and the people of Rome despaired greatly in the year of Judaea (or of the Jews), so O Lt, may there be protection from misfortune”.



Abbadi, S. 2001. A New Safaitic Inscription Dated to 12 –9 BC. In: Studies in the History and Archaeology of Jordan 7: 481-484.

Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (SSLL 80). Leiden: Brill.

Marginal Notes on: Written in Stone, the Smithsonian exhibition of Safaitic inscriptions

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions

The Smithsonian National Museum of Natural History website, in conjunction with the Saudi Ministry of Education, has published an online exhibition of inscriptions from the national museum of Saudi Arabia. This commendable effort, however, is undermined by the extremely poor presentation of these texts, often with completely unintelligible readings and translations. To illustrate, take the reading and translation of the Safaitic inscription 9:

L’qadam bin Lamo w hadafa hal lat slm
Proposed Translation
The person asking favor from the deity E’lat
The message asks a favor from the deity E’lat. “E’lat” is the ancient name of a god in northern Arabia. The person is asking to be “salam” (meaning “alright,” “good,” or “at peace”).

I have emailed the director of this website twice offering my expertise to improve the exhibition, as this is one of the first hits one gets when searching for “Safaitic” or “Ancient North Arabian” online. Neither email received a response. I have therefore decided to correct the readings and interpretations of these texts in my Marginal Notes series. Unfortunately, the locations in which these texts were discovered are not provided.


Safaitic 1


Revised reading and translation

l bgt bn gdy bn ls2ms1 w nṣb w…

By Bgt son of Gdy son of Ls2ms1and he erected a cult-stone.

Notes: On the ancient Near Eastern tradition of erecting a stone in commemoration of a religious event, see Macdonald (2012).  The remainder of the inscription is broken off.


Safaitic 2


Revised reading and translation

l {ġ}ṯ bn mʿz bn m{ġ}r w dṯʾw {d}s²r s¹lm w ḥ{l}{l}

By ġṯ son of Mʿz son of M{ġ}r and he spent the season of the later rains, so, O Ds²r, may he be secure and camp (?)

Notes: dṯʾ is one of the main seasons of the nomads of the Ḥarrah, stretching from mid-February to mid-April. Its beginning is signalled by the zodiac sign mlḥ ‘Aquarius’ (Al-Jallad 2016).  The final phrase w {d}s²r s¹lm w ḥ{l}{l} was read by Chiara Della Puppa.


Safaitic 3


Revised reading and translation

[l] Nẓrʾl bn ʿdy bn ḥddn bn nẓrʾl {w} qṣṣ ʿl- gm…w bzy ḥwlt w ʿqrb w {b}ṯ hfʿ{n}

[By] Nẓrʾl son of ʿdy son of Ḥddn son of Nẓrʾl and he patrolled on the border of Gm{.} and he subdued the Ḥawālat and ʿqrb and drove off Hfʿn

Notes: This interesting and rather non-formulaic inscription describes the military activities of its author. The verb bzy attested for the first time here is likely cognate with Classical Arabic bazā-hu ‘he overcame, subdued him’.  Chiara Della Puppa suggests that the b be read as a broken ġ, which would mean ‘he raided’.  The normal form in Safaitic is ġzz but ġzy is rarely attested.  Nevertheless, the letter form resembles the b’s attested earlier in the inscriptions.  The Ḥwlt are a well known tribal group of North Arabia, attested as adversaries of the authors of the Safaitic inscriptions and the Nabataeans (Abbadi 2015). ʿqrb in this context must also refer to a social group. The next verb bṯ is probably equivalent to Classical Arabic baṯṯa ‘to disperse, scatter’. The following word, hfʿn is likely a social group, perhaps related to South Arabian hfʿm.


Safaitic 4


Revised reading and translation

l mryġṯ bn tyd bn ḥfry w dṯʾ

By Mryġṯ son of Tyd son of Ḥfry and he spent the season of the later rains (here)

Notes: The name Mryġṯ should probably be vocalized as Mar-Yaġūṯ, meaning ‘man of Yaghuth’. The latter element is the name of a pre-Islamic Arabian deity known from the Qur’an (Q 71:23).


Safaitic 5


Revised reading and translation

l ḫzn bn zbd bn ḫzn w s2ty

By Ḫzn son of Zbd son of Ḫzn and he spent the winter (here)

edit: See comment section.

Safaitic 6


Revised reading and translation

l ḥbb bn gs2m w wgm ʿl- ʾḥs1n w ʿl- ḏl

By Ḥbb son of Gs2m and he grieved for ʾḥs1n and for Ḏl


Safaitic 7


Revised reading and translation

[l] ġṯ bn klb bn brd w gls¹ … w ṣyw

[By] Gṯ son of Klb son of Brd and he halted …. and suffered from the lack of rain (?)

Notes: the verb ṣwy is well known and could mean ‘to build a cairn’ or ‘to suffer from the lack of rain’ (Al-Jallad 2015: 347). In the present inscription, the final two letters are metathesized. This could be a writing error or a peculiarity of the author’s dialect.


Safaitic 8


Revised reading and translation

l km bn km bn gfft bn mṣry bn ʿgr w bn{y}

By Km son of Km son of Gfft son of Mṣry son of ʿgr and he built (or lay a stone on a cairn).

Notes: The expression bny ʿl ‘built upon’ seems to refer to the tradition of laying a stone on the grave or burial cairn. The terse language of this inscription does not allow us to determine what sense is intended here.


Safaitic 9


Revised reading and translation

l qdm bn lṯ w wḥd f h lt s1lm

By Qdm son of Lṯ and he was alone so, O Lt, may he be secure.


Safaitic 10


Revised reading and translation

l ġlb bn ʾʿtl w ts2wq w ktm

By ġlb son of ʾʿtl and he felt longing but concealed (the object of his longing).

Notes: Most inscriptions containing the verb ts²wq ‘to long for’ also include the object of said longing, introduced by the preposition ʾl ‘for’. This individual chose instead to conceal the identity of the person he misses.



Abbadi, S. 2015. New evidence of a conflict between the Nabataeans and the Ḥwlt in a Safaitic inscription from Wadi Ram. Arabian Epigraphic Notes 1: 71-76.

Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (SSLL 80). Leiden: Brill.

Al-Jallad, A. 2016. An Ancient Arabian Zodiac: the constellations in the Safaitic inscriptions, Part II. AAE 27: 84–106

Macdonald, M.C.A. 2012. Goddesses, dancing girls or cheerleaders? Perceptions of the divine and the female form in the rock art of pre-Islamic North Arabia. Pages 261–297 in I. Sachet & Ch. J. Robin (eds), Dieux et déesses d’Arabie Images et représentations. (Orient et Méditerranée, 7). Paris: De Boccard.


Marginal Notes on: “Safaitic inscriptions from the Eastern Part of Mafraq Governorate/ Jordan”. By Sultan al-Maani and Fardous al-Ajlouny, in Adumatu 8 (July 2003), pp. 33-49

This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions

This article publishes a relatively long Safaitic inscription (no. 4), which the editors read and translate as follows:

[l] s²qq bn s²hyt bn ʾs¹ bn ḥg bn s²bḥr bn grmʾl bn ʿbṭ bn ʿzhm bn mrʾ bn ʿrs¹ bn rġm bn s²hr bn rṭḫ bn ʿwḏ bn whbʾl w byt b- ʾbl w wrd l- ġdf f h bʿls¹mn rwḥ w s¹lm w mḥlt l- ḏ yʿwr h- s¹fr

‘By PN. He lodged his camels for the night and came to the water (the filthy water or he came to the water at night). I pray for Baʿalsamīn for relief and peace. God may bring sterility to those who defect this inscription’.

Photograph of MH4 (
Photograph of the inscription (from original article)

Suggested reading and translation:

The reading is sound with the exception of the name rġm, which should be read instead as rġs¹. The initial lām auctoris is clear and need not be in brackets.

New translation: ‘By PN and he camped the night with camels (or among camels) and went to water seeking abundance so, O Bʿls¹mn, send the winds so that he may be secure; and may he who would efface this writing experience a dearth of pasture.’

wrd l-ġdf: The edition took ġdf as meaning ‘night’, based on the example ġadafa l-laylu ‘the night was lowered’ in Ibn Manẓūr. The meaning ‘night’, however, is not part of the verb. It seems better to take it as a noun related to Classical Arabic ġadafun ‘abundance, ease’, and the preposition l- here indicating purpose. The verb wrd does not take a locative goal with the preposition l-, for example wrd h-bʾr b-h-nmrt ‘he went to water at the well near Namārah’ (Al-Jallad 2015: 227; RSIS 339).

mḥlt: This common noun is better translated as ‘dearth of pasture’; (Al-Jallad 2015: 327)


The inscription is a prayer for rain written by a man seeking pasture for his camels. It seems that the rains were poor and there was not sufficient herbage in the expected areas.  Therefore, the writer was returning to a place with permanent or seasonal water (wrd). His prayer to the storm god, Bʿls¹mn, for rain (rwḥ) and the curse of dearth (mlḥt) upon those who efface fit this theme as well.


Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (SSLL 80). Leiden: Brill.


This belongs to a series of “marginal notes” I will post on recently published Safaitic inscriptions

The present edition of inscriptions was published without photographs or tracings so it is impossible to verify the readings from the rock. Nevertheless, two of the inscriptions can be read in a more satisfying way, and these can potentially shed light on a new shade of meaning for the word ṣrt ‘enclosure’.

Inscription 2

Ed. pro.: l bnʿtm bn qym twlh ṣrt

‘By PN (and he) deeply grieved and cried/afflict of grief’


Suggested reading and translation:

l bnʿtm bn qymt w l-h [h-]ṣrt

‘By PN and [the] enclosure is his’

twlh > w l-h: The edition took the four letters following the name qym as a T-stem verb derived from the root wlh ‘to grieve’, Classical Arabic waliha. This would be strange as the narrative is almost always connected to the name with a conjunction. It is more likely that the t belongs to the previous name, which the addition acknowledges but does not give as their primary reading. Thus, the name of the father would be qymt, which is well attested in Safaitic and found in Greek transcription as Καιαμαθος /qayyāmat/. The narrative would then give the common ‘building’ or ‘ownership’ formula, w l-h [h-]ṣrt.  The spelling of the clitic pronoun and following definite article with one h is common (Al-Jallad 2015:49).


Inscription 4:

Ed. pro.: w ndm ʿl- bnʿtm wlh ṣrt

‘and he anguished for Bnʿtm, (he) grieved and cried strongly’


Suggested translation:

w ndm ʿl- bnʿtm w l-h [h-]ṣrt

‘and he was devastated by grief for Bnʿtm, and the enclosure is his’

The referent of the clitic pronoun is unclear; it could be Bnʿtm or it could refer to the author, who has now claimed the enclosure.


Remarks on the meaning of ṣrt

Inscription 4 is a funerary inscription and so it could be argued that ṣrt ‘enclosure’ here functioned in a similar way to rgm, ‘funerary cairn’ (Al-Jallad 2015: 337). This would moreover suggest that inscription 2 was a burial text, so: for Bnʿtm son of Qymt is the burial (?) enclosure. The other texts published in this article contain expressions of grief for Bnʿtm, similar in fact to the types of rituals associated with the rgm and ṣwy, e.g. bny ʿl- Bnʿtm lit. ‘he built over/for Bnʿtm’


Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (SSLL 80). Leiden: Brill.

A few notes on the newly discovered Hasaitic-Aramaic bilingual from Mleiha in the UAE


Many thanks to the The British Foundation for the Study of Arabia for sharing a photo of the partial bilingual Aramaic-Hasaitic inscription from Mleiha (UAE) (see the following popularizing article:’s-tomb-at-Mleiha-in-UAE’s-Sharjah-emirate). From what can be read on the photograph, which is not of the highest quality, the Hasaitic text states:

nfs w- qbr ʿmd bn
gr bn ‘ly bqr mlk
‘mn ḏy bn…. br
-h ‘md bn …. d
bqr m…

The memorial and grave of ʿmd son of Gr son of ʿly bqr of the king of ‘mn, which his son ‘md son of …. bqr of the [king] constructed [for him]…

The text is composed in the Ancient South Arabian (Musnad) alphabet, in a particular variety used in eastern Arabia known to scholars as Hasaitic.  The inscription is not particularly informative from a linguistic point of view, but nothing contradicts its interpretation as Sabaic, with the exception of perhaps an important linguistic deviation, the word for “son” in line 3. The name of the deceased is spelled as ‘md bn gr bn ‘ly, where bn is the normal word for ‘son’ in Ancient South Arabian, and indeed in most Semitic languages. However, in line 3, the building formula uses the word br for “son”. Br (probably pronounced bar) is the normal word for son in Aramaic, but why an Aramaic intrusion here and not in the genealogy? It is possible that building formula was taken directly from the Aramaic portion of the text, but written in Ancient South Arabian letters. The inscription is damaged following the second letter of the word “built”, so it is impossible to comment on the spelling. One objection to this is that the word “which” ḏy agrees with the Ancient South Arabian form, namely, a dual referring back to nfs ‘memorial’ and qbr ‘grave’. In general, however, the Aramaic text is shorter and seems dependent on the Ancient South Arabian one, so the importing of an entire clause from the former seems unlikely.

Another, more tantalizing explanation is possible. The use of br may reflect the spoken language of the inhabitants of ancient Mleiha. While bn is used in the genealogies, the common formula no doubt assured the use of the correct Ancient South Arabian form. However, in the building portion, the scribe may have subconsciously used a word from his spoken language. Aramaic was surely foreign to the area, so unless we imagine a colony of northern immigrants, Aramaic seems like an unlikely source. There is one other group of Semitic languages that uses the word br for son, the so-called Modern South Arabian languages, such as Mehri, Shehri, Harsusi, etc. which are still spoken in Oman today, yet not this far east. The word br for son is attested, as far as I know, in one early Sabaic inscription (Y.85.AQ/11 Eryani 57) from Shiʿb al-ʿAql near al-Jafnah ( Could it be that in former times the ancestor of the Modern South Arabian languages was spoken as far east as Mleiha and as far west as Dhamar, but receded under the spread of the Ancient South Arabian languages first, and ultimately Arabic? It is an exciting possibility that must await further discoveries to confirm. Nevertheless, each new epigraphic find will shed important light on the rich linguistic landscape of pre-Islamic Arabia.