Tuesday, August 28, 2012

A Ramesside North Arabian (?) Inscription in Early North Arabian (?) Script

A facsimile of the drawing of this inscription can be found here: http://www.britishmuseum.org/pdf/1c%20Hieratic%20inscriptions.pdf - Figure 6; on page 26 (bottom).

The inscription should be read from left-to-right.  Before offering a translation, it is important to note three non-linguistic (or quasi-linguistic) paleographic features: 1) immediately before the inscription is a drawing of a jackal (and possibly something by its tale); 2) immediately after the inscription is something else (unclear, possibly a star); 3) above the inscription 'ankh-w3s' is drawn.  There is also a sort of a cross-like mark beneath the inscription.

The inscription should be read as follows:

hṣ . qṭ . ḏypm  /  Haṣu . qaṭû (?) . ḏīpam


"Has (PN) approached the Jackal."

So the inscription itself offers some subtle linguistic clues that in turn probably validate my interpretation of the paleography as generally a variant form of North Arabian.  The -m probably acts as a form of locative or accusative case-marker here.  It almost certainly confirms that the language was inflected.  Additionally, the preservation of ḏ in ḏyp (root ḏyb) suggests the dialect was not subject to Proto-Canaanite or Aramaic consonant mergers.  However, the /p/ is unusual if not unique, and despite the Egyptian cognate z3b may have been an attempt to equate Semitic /p/ with Egyptian /b/ (if 'The Jackal' refers to an Egyptian divine symbol).  Additionally, the ḏ is really only paleographically related to North Arabian scripts, in a circuitous way at best.

However, the most difficult aspect of this inscription is not ḏypm - which clearly relates to the picture of the jackal; it is the first four letters.  If one retrodicts that a verb must be present as a result of the -m locative, then either the verb is all four letters (implausible), is causative (h-), or consists of S-V or V-S.  In this case, ṣqṭ is almost certainly meaningless (although ṣqṭn appears once in an unclear context in a Sabaic inscription). This leaves the very unusual hṣ and qṭ as the most likely parsing of the words.  But problematically, the -m on the last word does not actually help frame the grammar.

However, it is possible that in Safaitic a name hṣ did exist (see: hṣ in KRS 837; hṣy in KRS 1884); in South Arabian it is possible a single Minaic inscription attests the name outright, but one might trace it also to a Sabaic group-name compound - hṣn`m.  Moreover, qṭ may be related to a fairly common verb found in Akkadian (a Neo-Babylonian word qaṭû that CAD calls it an Aramaic loanword - see qṭy in Syriac) meaning 'to near.'  The word (qṭw) is translated 'walking' (noun) in a single Sabaic inscription (Jamme 2870).

However, it is difficult to discount the Arabic qṭ "to cut" in the D-stem meaning 'to carve' (i.e. "Has carved the Jackal.").  But this derivation, possibly found in the Neo-Babylonian qāṭû (CAD also calls this an Aramaic loan), might then be related to Aramaic qṭw meaning 'cane.'  If connected, the earlier meaning apparently refers specifically to cutting wood or reeds - and so probably is not relevant to this inscription.

If this is a statement related to death and Anubis, it may make sense.  The odd use of /p/ finds parallel in Thebes 2 (forthcoming?) in which bṯn (in this context a constellation meaning 'serpent') appears to be approximated pṯn.  But moreover, paleographically that /p/ (if correct), can really only be compared with Thamudic B, C, and D and Hismaic.  And the ḏ can really only be compared with Dedanitic (sometimes also called Lihyanite (in the later phases of development)).

So based on the paleographic and probable lexical triangulation, the actual language (reconstructed solely through these three words) appears to be a variant Syro-Arabian dialect.  The name is found in Safaitic (North Arabian) and Sabaic (South Arabian); the verb is found in Syriac, Neo-Babylonian, and Sabaic; and the 'jackal' is common but the orthography here appears to be unique.

The other really interesting thing is the apparent paleographic retention of a human with one arm raised and one lowered for /h/.  This find direct parallel (so far) only at Wadi el-Hol, although Jamme 863 uses a slightly more de-styled character.

Friday, August 10, 2012


This is not my best-researched effort (which likely says a lot).  According to Wikipedia, which provides a transliteration, Cyrus Gordon published this inscription (and transliteration) in Evidence for the Minoan Language (1966).  I have been unable to find serious subsequent decipherment attempts - although I haven't in earnest started amassing evidence for a literature review.  Taking the account at face value - and in light of the bilinguality of this inscription from Amathus, I would like to offer my own explanation of the inscription as either West Semitic or specifically Canaanite (this depends on the interpretation of the vowel /o/ as an aleph (i.e. Canaanite) or just another transcription of /u/ - which is the basis of my first attempt):




Cyrus Gordon translates this text as

The city of the Amathusans (honored) the noble Ariston (son) of Aristonax."

So in light of that Greek inscription/translation (I have reformatted/re-parsed the syllabic transliteration but I have also taken it from Wikipedia):

1: a-na ma-to-ri U-mi E-s[a]-i mu-ku-la-i . la sa-na A-ri-si-to-no se A-ra-to-wa-na-ka so-ko-o . se

2: ke-ra-ke-re tu-lo . se ta-ka-na-[?]-so-ti a-lo-ka . i-li-po-ti

Ana matori Umi Esa'i mukula'i . La sana Arisitono se Ara(s)towanaka sokō . se

keraker tulo . se takana[?]soti aloka . ili-poti

"To the small boat of Um-Hasa the mooring point, for the Second Ariston of the regal Ara(s)towanaka.  This is the

'hanging talent' [of gold].  This is my tribute to you.  Ili-Puti."

Then Amathus would actually be a corruption of Um-Hasa; really probably Um-Hatha.  The occurrences of /o/ are interesting - mator (Akkadian maturru), Aristono (Greek Ariston); Aratowanaka (Greek Aristonax), soko'o (Akkadian sukkû from Elamite sunki-), tulo (Akkadian tullû); Ili-Poti (probably Ili-Puti (the second part of the compound frequently found with gods in Akkadian)).  The only outlying orthography would then be aloka (which I take as a form of 'lyk - found in Hebrew, Aramaic, and Arabic) - and the reason for this vowel may simply be a vowel interchange issue in which /ei/ interchanged with /o/ so that 'aleyka becomes aloka.

So two possible conclusions: The dialect 'heard' and approximated /o/ from (Standard? or Neo-?)Babylonian u and from Greek o.  It is also possible (if not likely) that the first sentence's inflection is anomalous and the dialect also did not possess inflection.  The first sentence, complete with Akkadian preposition, would then be an attempt to employ a very formal (archaic) language, namely (probably) Akkadian.  And the balance of the text simply ignores this inflection.

Nobody's ever actually posted a question, but preemptively, my interpretation of mukula comes from the Akkadian makallû a Standard and Neo Babylonian word for "mooring place."  Exactly why this would be the chosen term to describe Amathus is unclear.  It could, however, have been the dialect's chosen word for a coastal city, inasmuch as the juxtaposed Greek word is polis.  However, CAD connects the verb kalû really 'to detain.'  I wonder if it isn't a related but slightly different term for coastal (or non-river) city from mu 'water' and kula 'detained' - possibly meaning either 'shallow water' (i.e. mooring place), or a place 'denied water' (kalû can specifically mean 'to deny irrigation water').

Sunday, June 24, 2012

The Deir Alla Corpus

(This is from my comment (below) regarding the third (astronomically-themed (?) inscription):

Finally, cNT-NR might not be taken as a reference to Anat, but rather perhaps as 'furrow' as in Ugaritic - with recognition that this dialect is not, strictly, formalized Aramaic.

Then the entire translation of III would read:

“He is Nasi-Salmu (My Hawk is Salmu), the Chief of Saran; the Raiser(?)-of-Fire
Has softened the Furrows-of-Fire; the Hook-of-Fire has been hardened
By the Light.”)

This is a relatively short paper hopefully for submission to BASOR on the tiny Deir Alla corpus of three inscribed clay seals, two of which are dotted.  It is unfinished as can be seen in my use of XXX instead of citations in a few cases.

Hatarūnī, a Dialect of Northwest Semitic:
New Readings and Translations of Inscribed Alphabetic Clay Seals from Deir cAlla

This is a new proposal to read three inscribed clay tablets from Deir cAlla dating from before 1200 BC.  They occur as several generations of abjadic writing, known so far only in these three inscriptions.  There are essentially two substrate script trends, both of which are relatively unique.  The earlier is distantly related to late Proto-Sinaitic, but also offers clues to the origination of the later North Arabian scripts.  The later exhibits paleographic features suggesting increased contact of Proto-Canaanite.  Of particular interest are likely recordings of a solar eclipse, and the expansion of the range of attested Northwest Semitic dialects.


This is a new proposal to read three inscribed clay tablets from Deir cAlla (Deir Alla) as Northwest Semitic inscriptions, distinctly influenced by Northeast Semitic.  The later relatives of this script are attested, at present, only much later beginning in the late 2nd millennium BC at the earliest, but really in the late centuries BC and early centuries AD.[1]  Moreover the languages of those later scripts are not closely related to the languages of the Deir Alla inscriptions.
Thus these inscriptions may evidence a much earlier incubating paleographic trend, not currently well-known, that impacted the development of later North Arabian scripts.  The texts expand the attested range of Northwest Semitic.  The name ‘Hatarūn’ occurs apparently in reference to the probable intended recipient of Document I, and thus the gentillic adjective would be Hatarūnī – possibly representing the name of the people related to these texts.
The only other attempt to decipher these inscriptions was conducted by Dr. William Shea.[2]  While at the time the attempt was reasonable, in light of new evidence and a broader re-interpretation of the early alphabetic evolution, the paleographic assessment is inappropriate.  As a result, the epigraphic analysis is also no longer tenable.  The paleography of the Deir Alla scripts relates more closely to North Arabian, even if epigraphic concerns veer off in a different direction.  The decipherment provided here – while philologically unusual – offers insight into two (probably progressive) marginal dialects of Northwest Semitic, with archaic features and Northeast Semitic influence.
The paleographic re-identifications here involve a much larger project to classify and decipher the entire early alphabetic corpuses.[3]  While the exact placement of the Deir Alla scripts within the early alphabetic dialect spectrum is debatable; they are connected to a complex evolutionary spectrum of the early alphabet – stretching from at least 1800 to probably 1200 BC, after which most variant corpuses were subsumed essentially by the more formal Proto-Canaanite, and North and South Arabian corpuses.
While Shea’s analysis of these tablets may seem logical out of this very broad context, the differentiation in the orientation of T and Ḫ in I strongly suggests a tendency toward North Arabian paleography.  Moreover, the repetition of a word [rs1; Shea’s mk], and the quite unmistakable presence of the grapheme S1 necessitates the consideration of this small corpus as paleographically Syro-Arabian; Proto-Canaanite never possessed it.  While it appears infrequently in Proto-Sinaitic, and in Wadi el-Hol, it primarily occurs in in North and South Arabian –in Ugaritic shibbolethed from š to s.[4]
Substantial evidence in the broader study of early alphabets suggests a missing genetic link, which precedes Proto-Canaanite and even possibly early Proto-Sinaitic, between the Wadi el-Hol inscriptions and the later Arabian scripts.  Deir Alla I & II appear to form a fringe trend derived from that missing link.  But while they probably represent a derivation from a common ancestor of some North Arabian scripts, the early Deir Alla script are not that ancestor but is rather a fringe descendant.  Paleographic developments in III, likely the latest Deir Alla generation, again suggest this people’s writing system was subsumed in the formalization of Proto-Canaanite even before the destruction of their palace by an earthquake shortly after 1200 BC.[5]

A Variant Abjad Subsumed
The texts will be treated first with respect to their consonantal nature.  The unusual nature of the odd dots[6] in I and II will be examined in the following section, in which translations of these two texts will also be offered.  Because III does not obviously possess the dots of the other two, it will be translated in this section.

I:[7]  MKR / RS1 / HTRN / HBR / RS1 / LḪM

The context of this inscription is fairly clear.  The second name, LḪM, is almost certainly related to byt lḫm or Bethlehem.  But it suggests that the etymology of Bethlehem may indeed be from the deity Lakhmu, as early accepted by Albright.[8]  The name HTRN might be connected with the root ytr, related to “Aram /yatr/ rope; bowstring; Syr /ytr/ remain, be left over; be better, go ahead, increase, overhang, profit… Amor /ytr/, /wtr/ be more, abundant, excellent; … Arab… be stretched”.[9]  Despite Murtonen’s Akkadian association, the above can be connected with Akkadian itartu “additional pack (of a donkey load); OA” [10] pertaining to atāru (watāru) “to exceed”.[11]
But HTRN, probably a name, may be a C-stem.[12]  Moreover, the plural found probably precludes its (non-Aramaic) Northwest Semitic origin, thus relating it to either Syriac ‘helpers’ or ‘those who have profited’,[13] or Arabic ‘stringers (of the bow)’.[14]  MKR should also be treated as a noun, either a name, or ‘merchant,’ as with “Phoen /mkr/ sell; seller, dealer, merchant; Ug /mkr/ merchant; Aram /mkr/ buy; promise (one’s daughter) in marriage… Syr /mkr/ buy; (be) betroth(ed)”,[15] and with Akkadian makkāru[16] and mākiru[17] both “trader.”
The remaining word HBR is unusual, but within the context of an apparently Akkadian-inspired name Laḫmu, this word can be compared (phonetically) with Akkadian ibru “(ebru) a.; person of the same status or profession, comrade, fellow, colleague, friend; from OA, OB on”.[18]  There is little evidence for West Semitic confusion between H and Ḥ, so the orthography here may in fact reflect an Akkadian phonetic realization.  If the scribe were East Semitic, they may have been unsuccessfully attempting to approximate the West Semitic sound (i.e., H/ ~ Akk i/e).  Unfortunately the lack of an attested differentiated Ḥ in these inscriptions makes speculation difficult.
For rs1, present in each inscription, one can compare r’aš ‘head, chief’.[19]  But an explanation will be provided for the aberrant orthography at the end of this paper.  The reason for holding off on explanation is the connection (in I and II) of the dots and r’s, which corroborates the interpretation of rs1 but must be explained in the broader context of the unique system of marking matres lectionis.  The ubiquity of the phrase in all three inscriptions makes its interpretation relatively uncontroversial even without the phonetic explanation for the contraction of the vowel.

II:[20]  W[ṢYH] . NH K PG / RS1  |  WDH NḪ / LḪBR / [RS1][21]

This text is a boustrophedon.  The word WṢY can be related to Akkadian aṣû (waṣû, waṣā’û, uṣā’û, uṣu) “to leave, to depart”,[22] and the orthography in II also pertains to Ugaritic and Amorite.[23]  Here WṢY is apparently a noun, indicated by the locative –h, which is essentially a NW Semitic feature.[24]  The following word NH is common, meaning ‘sigh, shout’[25] and in Arabic really ‘mourn’.[26]  As in I, there may be Akkadian influence, which phonetically renders the verb na’û (nu’û) “to shout”.[27]  One could also see this as a Canaanite orthography, however.[28]  The G in this case, if correct, is one of only two known examples of this variant grapheme.[29]  Then (k-) pg might be compared with Akkadian pagû “a stringed instrument”,[30] or pigû of similar meaning.[31]
The next line is substantially clearer.  The second use of the locative –h in DH parallels this usage in the first line.[32]  The following NḪ reflects a noun form of ‘rest’ stemming from a fairly common root.[33]  The apparent recipient of the dedication, ḪBR, has an interesting name.

HW NṢ / ṢLM / RS1-S3RN / ’NS2-NR
RKK / cNT-NR / cRN / WW-NR

“He is the Dark Hawk, the Chief of Śaran; the Compeller-of-Fire
Has softened Anat-of-Fire; the Hook-of-Fire has been hardened
By the Light.”

This inscription is interesting for two reasons: it is essentially undamaged; and it appears to be an early recording of a solar eclipse and the first historical recording of Baily’s Beads – dots of light that appear and disappear along the side of the solar disc during a solar eclipse.  It is also not as obviously inflected as the previous two, and may exhibit dialect shift – toward a dialect closer to Syriac – but with archaic features.  The name Śaran is not clear – it could refer to the Plains of Sharon, but ‘(bn) s\rn[35] is a name attested in Ugaritic.  The word NṢ refers to a bird of prey, in Syriac apparently “hawk”.[36]  Conversely, ṢLM is almost certainly a borrowing from Akkadian ṣalmu, meaning ‘dark,’[37] also noting that ṣallamtu is the name of a bird.[38]  The word ’ns can be compared in Syriac with “to press, compel… to force”.[39]  If any of these epithets hide a name for the ‘Chief of Śaran,’ that name is not immediately obvious – unless perhaps nṣ ṣlm.
The next line evidences clear juxtaposition that supports a Syro-Arabian interpretation.  The verb RKK, is masculine and thus must take the following feminine noun as an object.  A D-stem in Syriac would mean “to soften, mollify; to make tender, gentle”.[40]  In contrast, cRN occurs only in Gt-stem form in Syriac as “to become hard”.[41]  However, the Arabic noun caran primarily means “a bony outgrowth, exotosis (med.)”.[42]  Though only in noun form, it argues for the theoretical existence of a G-stem at some point.  Moreover, the c grapheme[43] in crn is oriented differently than the c in cnt, suggesting differentiation of matres lectionis.  Since can(a)t likely has a historically stable vowel pattern, crn reflects cirn or curn.  As parallel grammar suggests crn is a perfective verb, the latter must be correct – in turn indicating the passive perfective curin.  This is probably unrelated to the ‘dots vowel system’ but rather an ad hoc attempt to distinguish the passive verb form.
Interestingly, the R in cRN evidences paleographic recidivism, based on the apparent continued usage of the letter name riš (possibly an earlier ruš – specifically at Deir Alla).  The scribe used an ad hoc but essentially obvious graphical derivation, deviating from the styled V-like-R to a head-like-R.  But this derivation also implies a regular R grapheme rotated 90° counterclockwise protruding from the ‘head,’ possibly as its ‘hair.’  Within the context of the inscription’s narrative, this ad hoc grapheme may have better conveyed the idea of Baily’s Beads alongside the eclipsed sun.
The two phrases in this line are interesting.  The first, cNT-NR, refers to the goddess Anat with an apparently unattested epithet ‘of Fire’.  Likewise, WW-NR appears to refer to a ‘hook of fire’ – with neither necessarily being localizable to a specific dialect.  The final line, written on the edge of the seal, has apparently lost an explicit aleph present in the root.  The reason is almost certainly phonetic.  The only independent aleph suggested is in III, and is paleographically very similar to H.[44]  In B[’]WR it seems most likely that no glottal was explicitly pronounced, i.e. būr.  The use of this prepositional phrase confirms the passive form of cRN to which it is connected.
            There is an additional point of confusion in translating this inscription in particular.  The above string of phrases reveal III’s pattern of placing construct-states within word divisions.  But the first division – hw nṣ then might be taken as (the odd) ‘he of the hawk.’  The following word ṣlm would then either reflect a D-stem verb ‘to darken,’ or a noun ‘dark one.’  While within the context of the other two inscriptions rs1 can reasonably be taken to mean ‘chief,’ it can also reflect certain geographical terms in both Aramaic and Arabic.  But, with obvious regard to context, two nouns in a word division, may also have been a means of rendering a nominal sentence in III.  The adjective ṣlm would then occur after the word division to avoid genitive confusion.

The Nature of the Dots
Quite interestingly, the dots littered throughout I and II are posited to reflect a short u, essentially abnormally early evidence of the ad hoc indication of matres lectionis.  The hypothesis is very simple – the dots were used to distinguish u from a, i, or no vowel.  This is not efficient as a diacritic system, so its origin may be in differentiation of grammatical possibilities – for which, along with word dividers, it is surprisingly efficient.  This system is hypothesized as an ad hoc development based on a total lack of additional external evidence for a comparable means of differentiating matres lectionis, inflection, or grammar.
In III, construct states appear to occur only within word divisions.  This may be a late feature.  In I and II construct states appear to straddle word divisions, with the second noun not being marked with a dot (i.e. non-nominal case).  When placed above or beside a letter, the dot may reflect u.  But in rs1, it may reflect a condensed glottal ’u to /u/.  Two dots may specifically reflect ū in the plural htr[uu]n – but this plural is not used grammatically in context (it is a name).  It is likely that the loss of dots in III indicates or follows a progressive loss of inflection.  It is not clear whether the dots in I and II may inadvertently offer early evidence of the Canaanite shift from ā to ō.
What must have been an ad hoc system could be attributed to a single scribe.  By III, there had been at least three generations of script developments; but there may well have been more.  The first two are epigraphically and paleographically more similar, but II may be the most archaic of the three.  The eclipse text III must be from a significantly later period.  The script may then have been transmitted to the scribe of III, probably interpersonally, without the ad hoc vowel system – or that scribe chose to stop using it.  This fits a gradual replacement of the unusual and sparsely attested script by scripts related to Proto-Canaanite.
The -*, next to a letter, reflects being written on top of (or connected to the top of) the letter to the left; when written with a space, it reflects a dot drawn next to the preceding letter but not obviously ‘connected’ to it.

I:  MKR* / R*S1 * / HTR* *N / HBR* / R*S1 * / LḪM
makaru . rušu . hatarūni . habru . rušu . laḫami
“Makaru (/the merchant?), Chief of the Hatarūn, is an ally.  Chief of Laḫam."

II:  WṢ(*?)YH N*H KPG / R*S1  |  W DH NḪ / LḪBR* / [R*S1]
waṣuyah . nuha . ka pagi . ruša  / wa . dah . niḫa . li-ḫabiru . ruši
“At the departure is lamentation, like the pagû-instrument, to the chief –
And at this (time) is rest for Ḫabiru-the-Chief.’

It is possible that I is some sort of travel document pertinent to a nomadic people.  A similar idea for diplomatic agents is known during the Amarna Period.[45]  If this is indeed ‘permission’ from the ‘Chief of Bethlehem’ for passage or trade through the area, it would suggest the more widespread attestation and usage of this script, for however brief a time.  It is important to note that the proper understanding of ibrû is “an institutionalized relationship between free persons of the same status or profession which entailed acceptance of the same code of behavior and an obligation of mutual assistance.”[46]  It is also possible this essentially codifies an alliance.

It is probably also worth pointing out how brilliantly II is written in a sort of phonetic juxtaposition between the two lines: wa1-wa2, (ṣuya1-da2,) ah1-ah2, nuh1-niḫ2, (li2,) ka1-ḫa2, pa1-bi2, gi1-ru2, ru1-ru2, ša1-ši2.  Considering the parallel use of the locative/directive cases, the contrast of a root w- and the conjunction, the parallel use of prepositional phrases, and the lines both ending with the same word inflected differently, this may be considered a poetic composition – certainly in a funerary context.

Thoughts on Paleography
In I and II, Ḫ can really only be compared paleographically with the North Arabian Thamudic B&C, Hismaic, and Safaitic trends.[47]  In I, this grapheme can be clearly differentiated from T.  Unfortunately, no examples of a distinct Ḥ are attested so far in this script, which would distinctively mark it as being Arabian.  But in addition to W, the M’s present in I and III are also distinctly Arabian – enclosed forms.  These must be compared with Taymanitic, Thamudic B, Safaitic, or Hasaitic, or perhaps Hismaic,[48] in addition to South Arabian.
In general, it would be fairer to say that this script evidences some of the potential source of the later North and possibly South Arabian paleographic variation.  This manner of contact is not unbelievable as Deir Alla is at a crossroads of sorts.  However, the script attested in I and II cannot actually be ancestral to North Arabian.  Rather, these three short inscriptions provide evidence of the early alphabet’s ‘incubation,’ in various places, in turn explaining connected variance between earlier and later periods.  By III, the presence of a Phoenician-style and the essentially Canaanite s (S3) additionally support the conclusion of the early script’s gradual replacement by a formal corpus.

Epigraphic Thoughts
As for the language, ruš is an unusual orthography for ‘chief.’  However, the Amarna orthography rūšu ‘head’, written ru-šu in cuneiform,[49] offers a very solid comparison as possibly does a (foreign) name found in Ugaritic ’ilrš.[50]  The semantic connection of ‘head’ and ‘chief’ is common.  Two cases of h certainly correspond to Akkadian vowels: both habru and nuha appear to reflect Akkadian orthographies.  In II, li-ḫabiru / ruši reflects a prepositional phrase (i.e. liruši), but the dot after ḫabiru marks it as a name.  Likewise, the context and presence of the preposition negate the possibility of a construct state.[51]
In I, dots occur next to, rather than above, two S1’s.  In I and II, dots occur above two R’s but at the end of words (rather than initially as with ruš) – both in names.  It is possible that this is intended to differentiate between grammatical usage and usage in names.  It is interesting to note that aside from with habru (which may be explained by severely condensed spacing), the only two cases in which dots occur above letters at the end of a word is in the two PN’s present in I and II.
The unusual –a in the nouns following the locative cases is unclear.[52]  In II, however, ruša clearly reflects an accusative case without the addition of a preposition.  The origin of the plural form –ūn(a) is also unclear.  Though the language reflected is undoubtedly Northwest Semitic, it possesses archaic features, and was lexically and orthographically affected by contact with Northeast Semitic, and its contemporary cuneiform orthographies.

[1]  Michael C. A. Macdonald, “Chapter 6: Ancient North Arabian,” in The Languages of Syria-Palestine and Arabia, ed. Roger D Woodard (Cambridge, MA, 2008), 488-94.
[2]  William H. Shea, “The Inscribed Tablets from Tell Deir cAlla Part I.” Andrews University Seminary Studies, 27, no. 1 (1989a): 21-37; and William H. Shea, “The Inscribed Tablets from Tell Deir cAlla Part II.” Andrews University Seminary Studies, 27, no. 2 (1989b), 97-118.
[3]  For reference, I have included my charts as Appendix I; but even this is insufficient without the larger set of translations and external evidence.  And this is simply to show the distance between earlier and later alphabetic trends and the Deir Alla inscriptions.  Hopefully forthcoming JNES, XXX.
[4]  It additionally occurs in three of the ‘Theban texts’ published by Petrie (Thebes 1, Thebes 2a, and Thebes 6); see William Flinders Petrie, The Formation of the Alphabet, BSAE Studies Series, vol. III (London, 1912), Plate I.
[5]  H. J. Franken, “Clay Tablets from Deir ‘Alla Jordan.”  Vetus Testamentum, 14, no. 3 (1964a), 377.
[6]  It’s worth noting that Franken additionally published uinscribed but dotted seals, though these clearly did not serve the same purpose; see, H. J. Franken, “Excavations at Deir 'Allā, Season 1964: Preliminary Report.” Vetus Testamentum, 14, no. 4 (1964b): Plate Va.
[7]  I have been unable to find a photograph of this inscription, but see drawings in Franken 1964a, 379; and Shea 1989a: 29.
[8]  W. F. Albright, “The Canaanite God Ḥaurôn (Ḥôrôn).” The American Journal of Semitic Languages and Literatures, 53, no. 1 (1936): 7.
[9]  A. Murtonen, Volume 3, Sections Bb. Root System: Comparative Material and Discussions. Sections C, D and E: Numerals under 100, Pronouns and Particles, Hebrew Material, vol. 3 of Hebrew in its West Semitic Setting. A Comparative Survey of Non-Masoretic Hebrew Dialects and Traditions. Part 1. A Comparative Lexicon (Leiden, 1990), 225.
[10]  Chicago Assyrian Dictionary (CAD), I, 294, accessed May 22, 2012, http://oi.uchicago.edu/pdf/cad_i.pdf).
[11]  CAD, A: Part II:atāru, 487-92, accessed May 22, 2012, http://oi.uchicago.edu/pdf/cad_a2.pdf
[12]  The lack of an aleph fits this script’s tendency to drop alephs not at the beginning of phrases (for which there is only a single example).  In this case, the reason may be additionally phonetic as a C-stem would fit the realization ha’tar, condensed to hatar anyway without a distinct glottal-stop, and pluralized hatarūna.
[13]  Robert Payne Smith, A Compendious Syriac Dictionary: Founded upon the Thesaurus Syriacus of R. Payne Smith. (Oxford, 1903), 200.
[14]  ed. J. M. Cowan, Arabic-English Dictionary: The Hans Wehr Dictionary of Modern Written Arabic, fourth edition (Urbana, IL, 1994), 1227.
[15]  Murtonen 1990, 258.      
[16]  CAD M:makkāru, 131, accessed June 11, 2012, http://oi.uchicago.edu/pdf/cad_m1.pdf
[17]  ibid, 129.
[18]  CAD I:ibru, 5, accessed May 22, 2012, http://oi.uchicago.edu/pdf/cad_i.pdf.
[19]  Murtonen 1990, 391.
[20]  A photograph appears in Franken 1964b: Plate Vb.  Shea’s drawings occur in Shea 1989a: 31; and Shea 1989b: 99.
[21]  My reading of this word relies on Shea’s (1989a & 1989b) drawings.
[22]  CAD, A: Part II:aṣû, 356-85.
[23]  Murtonen 1990, 219.
[24]  Dennis Pardee, “Chapter 1: Ugaritic,” in The Languages of Syria-Palestine and Arabia, ed. Roger D. Woodard (Cambridge, MA, 2008), 26.
[25]  Murtonen 1990, 95.
[26]  ed. Cowan 1994, 1182.
[27]  CAD, N: Part II:na’û, 134, accessed May 22, 2012, http://oi.uchicago.edu/pdf/cad_n2.pdf
[28]  See, Ezekiel 7:11.
[29]  It occurs in the same constellation Theban text but in a relatively clear context (Petrie 1912, Plate I), the origin and descent of the odd anchor-shaped G is unclear.
[30]  CAD, P:pagû, 19, accessed May 22, 2012, http://oi.uchicago.edu/pdf/cad_p.pdf.
[31]  ibid,359.
[32]  Its consideration as a demonstrative is unlikely.  At least from a Proto-Sinaitic standpoint, the demonstrative’s orthography is always (> z), with no –h.
[33]  Murtonen 1990, 276.
[34]  A more than adequate photograph is provided; see Franken 1964a, Plate I.
[35]  J. L. Cunchillos, J. P. Vita and J. Á. Zamora, The Texts of the Ugaritic Data Bank, Volume II, translated by A. Lacadena and A. Castro (Piscataway, NJ, 2003), 995.  See also, J. L. Cunchillos, J. P. Vita and J. Á. Zamora, The Texts of the Ugaritic Data Bank, Volume III, translated by A. Lacadena and A. Castro (Piscataway, NJ, 2003), 1453.
[36]  Payne-Smith 1903, 347.
[37]  CAD, Ṣ:ṣalmu, 77-8, accessed October 4, 2011, http://oi.uchicago.edu/pdf/cad_s_tsade.pdf
[38]  ibid, 73.
[39]  Payne Smith 1903, 21.
[40]  ibid, 540.
[41]  ibid, 429.
[42]  ed. Cowan 1994, 712.
[43]  The Smithsonian Museum’s ‘Written in Stone’ exhibit’s Chart of Comparative Letters attributes very similar graphemes to ‘Thmaudic’ c’s (http://www.mnh.si.edu/epigraphy/e_pre-islamic/fig02_comparativechart.htm).  However, I am unable to verify this claim, and the credibility of this chart was called into question in my correspondence with another scholar of North Arabian.
[44]  This could suggest a possible parallel development to the origin, in my opinion, of Ugaritic ’i and ’u from the Ugaritic-Canaanite glottal H.  However, this may be a coincidence, and the aleph in III may simply be a development preceding some North Arabian variants employing two ‘hornlike’ lines protruding from either side of a straight line.
[45]  See, Bertrand Lafont, “International Relations in the Ancient Near East: The birth of a complete diplomatic system.”  Diplomacy & Statecraft, 12, no. 1 (2001): 46.
[46]  CAD, I:ibrû, 7.
[47]  Macdonald 2008, 496.
[48]  Macdonald 2008, 496.
[49]  CAD, R, 432, accessed May 22, 2012, http://oi.uchicago.edu/pdf/cad_r.pdf.
[50]  See, J. L. Cunchillos, J. P. Vita and J. Á. Zamora, The Texts of the Ugaritic Data Bank, Volume III, translated by A. Lacadena and A. Castro (Piscataway, NJ, 2003), 1453; the name ’ilrš occurs in several places – and may be a non-Ugaritic orthography meaning ‘Ilu is Chief,’ which would corroborate the analogous Deir Alla .
[51]  I suppose it is possible, due to the probable presence of ḫbr as a root in Ugaritic (KRT) (XXX), Akkadian (XXX), and Arabic (XXX) to interpret this phrase as ‘informant of the chief,’ though I believe this breaks the formula (in I and II) of PN [nominative] (Word Division) The Chief [nominative]; in this case the second nominative changes due to the preposition because it is not a proper noun.
[52]  Unfortunately an analysis of non-syllabic Ugaritic is not useful, and analysis of comparable Biblical Hebrew constructions is also inconclusive, not only because the directive occurs at the end of phrases but also because the loss of inflection renders analogy difficult.  The Deir Alla constructions are closer to that seen at Wadi el-Hol, in which the noun-directive heads a phrase.  The directive nature of the phrasing means that the following noun(s) really are objects – being directed toward the noun with the enclitic ‘case.’  This essentially flips the expected subject-object relationship, possibly indicating a passive construction: i.e. in Wadi el-Hol ḏunā ‘aṯtarah qašta kupṯa “These are for Athtar, the bow and the scimitar,” compared with li‘aṯtar qašta kupṯa “For Athtar are the bow and scimitar.”  The construction is almost certainly poetic in Deir Alla and may have been formal at Wadi el-Hol.

Thursday, May 10, 2012

The Wadi el-Hol Translation

I removed the previous versions of this paper because of a somewhat more specific approach I've taken (largely on philological grounds) over the past few weeks.  I will shortly post the paper in full.  I will also fix the presence of spell-check on the Wadi plates.

The primary distinctions (in the broader paper) result to an earlier structural sibilant/fricative usage present at Wadi el-Hol and in later Arabian script systems (but latent elsewhere), and the reinterpretation of a single character in Thebes 6, but the reinterpretation of dialect relations in some of the inscriptions as well.

With regards to Wadi el-Hol, my working hypothesis is currently that the dialect is closely related to Amorite.  This actually explains the structural sibilant confusion, directly reversed in Ugaritic; and the fricative (H) confusion is probably the result of a relatively early process of pharyngealization of H (although it is possible that there was a de-emphasis or glottalization of the pharyngeal).  In terms of graphemes, Ugaritic preserves all three 'H graphemes,' but reversed the usage of the pharyngeal and glottal.

Little regarding the translation or proper interpretation of Wadi el-Hol has changed.  I can now explain, however, why 'to intoxicate,' as a D-stem, lacks an explicit medial W.  I have also included vocalization with respect to the Northwest Semitic (Amorite-related) interpretation of verbs some of whose forms are best explained with respect to Akkadian (regardless of inflection patterns - which would then be NW Semitic).  This identification, along with some Proto-Siniaitic inscriptions will likely help clarify the apparent paucity of non-Canaanite Northwest Semitic languages (other than Ugaritic) through their more common attestation in the 2nd millennium BC, particularly in early alphabetic texts.

[From 'The Syro-Arabian, West Semitic Origins of the Alphabet in Egypt'] 
The Wadi el-Ḥôl Translation Hypothesis (1830-1800 BC South Egypt)
In 1999, two presumably alphabetic inscriptions were found on a desert trade road forty or so kilometers northwest of Luxor.  They have been dated tentatively to 1850-1700 BC.[i]  This translation is basically unexceptional, with a single Egyptian loanword.  There is some consensus between this study and previous identifications[ii] over the letters , B, , P, R,, and T.  Most previous readings omit a character before P in the Vertical Inscription.  This examination will thus concentrate on identifying the characters on which there is not the aforementioned agreement over identification:

Vertical: ?ṮTRH??T[?]PṮ’L
Horizontal: RB?N?NḤNP?H’??ḪR

Vertical Inscription:
The first character in the Vertical Inscription (see, Plate III) appears (partly) with only three-peaks in the Horizontal Inscription (see, Plate IV).  In the Vertical Inscription, these peaks are joined to a fourth, and the third is bisected.[iii]  Upon closer examination, the third peak is actually bisected by an N, and the three-peaked character is .  The fourth peak is then actually c.  A smaller, equally de-stylized wedge occurs above the following , below which is a line; there is a line also below the first character group (see, also Plate Va).  This then spells cṮTRH, the H possibly specifically being styled after Gardiner C10 – Ma’at.  The next character is a Q flattened on its left side, followed by S1, and T.  The inscription continues K, P, and, and ends with , L.

The text thus begins with a 3rd person dual demonstrative, ḎN.  In South Arabian ḏyn occupies this function,[iv] and a 3rd person dual demonstrative can be theoretically reconstructed in Akkadian as šunā.[v]  The non-feminized divine name cṮTR might reflect similarity with Ištar, probably present in Amorite,[vi] whose name could also connote ‘goddess’ in Akkadian.[vii]  Her name has also been traced to Proto-Semitic “*?aθtar- ‘morning/evening star’”.[viii]  This is followed by an enclitic locative/directive –h, found in Biblical Hebrew and Ugaritic,[ix] and meaning ‘for Aṯtar/the goddess.’  This grammar also explains the physical connection of Aṯtar’s name to the demonstrative.  The flattening of the left side of the Q may reflect it being styled as a “bow,” which is the word it begins QS1T (qaštu).[x]  A bow is in fact drawn (partly cut off in photographs) to the left of the depicted figure in the Vertical Inscription.

The next word, KPṮ may then be the staff to the right of that depicted figure.  It is probable that the K occurs above the T so that QS1T is not misread ‘archer,’ but rather that both are read together.  In Ugaritic, kpṯ occurs in an epithet to Anat.  Juxtaposed with “the high heavens”, some have argued kpṯ means “earth”.[xi]  It has also been linked with an Akkadian kub/pšu “headdress” and occurs among terms connoting authority.[xii]  The root kbʃ may also mean “Amor … lamb, young sheep; Akk … =; Arab … ram; chef; ram, pillar, buttress; Mhr … lamb; Soq ram…”,[xiii] though kpṯ may have reflected a different, if parallel, root at this time.  Indeed, the shape of the Egyptian xpS-scimitar has been described as the “foreleg of an ox”,[xiv] and Gardiner F23 (xpS) may also appear this way.  The Egyptian xpS (scimitar) has been connected with Greek κόπις (kopesh) of similar meaning.[xv]  The divine-royal object can now be connected with Anat’s Ugaritic epithet and this usage.  The last word is ’L,[xvi] meaning god or Ilu.

Horizontal Inscription
The bow and kopesh are known symbols, but the Vertical Inscription’s grammar disconnects Ilu from them.  The Vertical and Horizontal Inscriptions are paleographically unique, they are in the same location, and they probably come from a single time period.  They are also on the same rock spur – only several meters apart.[xvii]  These facts at least suggest the possibility that the inscriptions represent a single text.[xviii]  A line beneath the Vertical Inscription changes orientation at the end, maybe indicating the scribe ran out of space and continued writing several meters to the left (in the direction of the writing) on another inset surface of the same rock spur.  They were likely carved by the same scribe, and paleographic differences may partly result from fatigue or frustration in having to find a new spot and begin an even longer part of the text, possibly at night.[xix]

The first two words of the Horizontal Inscription (see, Plate IV) occur above a logogram (see, Plate Vb).  The first of these two words is RB,[xx] in Akkadian connoting “large… main, principle, chief”.[xxi]  But the logogram suggests a divine title known in Classical Arabic, and perhaps implicitly extant in analogy to the Ugaritic “rbt ’ṯrt ym … lady ’Athir(a)t of the sea”.[xxii]  The next word, DN should be compared with Amorite “/dnn/ be strong; /dunn/ strength”,[xxiii] also present in Amarna Canaanite and as Akkadian dannū “powerful”.[xxiv]  RB DN then means “Powerful Lord” in reference to the god of the Vertical Inscription.  Next, is a relative pronoun, i.e. “who,” like in Ancient North[xxv] and South Arabian,[xxvi] usually D in Ugaritic.[xxvii]

The following word, NḤ, is one of only four suggested cases in this paper of Semitic borrowing from Egyptian.  It may be connected with Egyptian nwḥ, “drunk,” reinforcing a connotation of Hathoric drunkenness.[xxviii]  The context and grammar demand a causative verb, and W may be drawn on the “arm” of (see, Plate Vc).  The connotation fits rituals to appease Hathor through “drunken, nocturnal revels… [to appease] the fiery power of the solar deity, an unquenchable flame to destroy the souls of the damned and a brilliant torch to illumine the paths of the blessed dead in the necropolis”.[xxix]    The offering of alcohol to the returning goddess Sekhmet was also probably not uncommon.[xxx]

The next word is NPS2, regardless of hesitancy over the last letter being S2.[xxxi]  This is not an improper orthography, but rather a variant Amorite orthography naps, alongside napš, “breath, life”.[xxxii]  The semantic development to ‘soul’ is not implausible.  The enclitic -H’, “her [soul],” and the locative case also mark this dialect as West Semitic.  Two unknown characters remain: W and an archaic , then finally and R.  Though perhaps not explicitly present in Amorite, the may reflect Ġ, thus reflecting the Babylonian-influenced orthography[xxxiii] of ṢĠR ‘(to be) small.’

At Ugarit, ṣġr is the name of a son of Baal and either Anat or a mysterious cow.[xxxiv]  There are thematic similarities between this story and the return of the oft-bovine Hathor to Abydos, resulting in the “conception, birth, and care of the child god”.[xxxv]  But the lack of stylization or logograms around ṢḪR, and its inclusion nondescriptly at the end of the inscription cast doubt on its divinity.  Akkadian ṣuḫḫuru means “to make smaller, to reduce (in size or number)”.[xxxvi]  NWḤ (‘drunk’) would have probably offered false(?) connections with the Semitic root NWḪ (‘rest’), in which case the former’s borrowed D-stem would be nuḥḥu, comparable to Akkadian nuḫḫu, explaining the written form.[xxxvii]

The lack of an explicit object for ṢḪR is explained by the unusual early presence of the conjunction W.  It serves to connect the subject-object relationship of NWḤ to the following verb ṢĠR; probably also suggesting parallel gemination.  It would appear that this scribe observed or participated in Hathoric celebrations.  It is thus not surprising that their inscriptions are on a path “ideally located to be integral in processions… welcoming… the returning goddess of the wandering eye of the sun”.[xxxviii]

ḏn  cṯtrh . qs1t kpṯ . ’l   Hrb dn . ḏ nwḥ . nps2h’ . w ṣḫr
Vḏunā caṯtarah . qaštu kupṯu / ’ilu . Hrabbu dannu . ḏa nuḥḥa . napsahā() . wa ṣuḫḫura

V“These are for Aṯtar (the goddess), the bow and the scimitar.  Ilu (the god)... H…[is] the Powerful Lord who intoxicates her soul and reduces it (of violence).”

Hence, the first inscription (part) depicts a warrior-goddess, the second a powerful god who intoxicates her, placating her violence of spirit.  Additional drawings, such as the bow and scimitar occur on both sides of the ‘figure’ drawn in the Vertical Inscription.  The stylization of the Q as a bow probably corresponds with the depiction of two scimitars to the left of KPṮ (see, Plate Vd), bounding the two words ‘bow and scimitar’ with two pictures of those symbols.  The Egyptian divine-determinative nṯr is certainly placed above Ilu’s name, and may unclearly be drawn next to Aṯtar’s name.

The Horizontal Inscription also evidences a probable simplification of the cuneiform divine determinative.  Hamilton speculated it may be a “subscript y?”.[xxxix]  However, there are two comparable cases (see, Plate Vb): in Sinai 365, and likely in the ‘South Arabian’ Jamme 863 where Albright and Jamme suggested it was a monogram of ḎT.[xl]  The Horizontal Inscription is also so far the only explicit evidence of an anthropomorphic origin for H.  But not least of all, this is the first actual text found in a dialect apparently related to Amorite – though not all of the vocabulary is known through Amorite names.

The Vertical inscription depicts neither an “ankh-sign”[xli] nor a woman.[xlii]  Though the figure does have a vaginal slit at its base, it is undoubtedly the goddess whose name appears directly below the figure itself.  Both inscriptions contain graphemes that stylistically might relate to nearby Egyptian prototypes from regnal years 26-30 of Amenemhat III.[xliii]  The content may relate to Middle Kingdom holidays[xliv] and to a Hathoric myth perhaps originating in the XIIIth Dynasty.[xlv]  They are most likely from 1830-1800 BC, well-preceding the Lachish Dagger of possibly 1750-1700 BC.[xlvi]

Addendum, A Sphinx Shaped Rosetta Stone:
I feel I should also post Sinai 345, as I slightly misinterpreted that inscription in my earlier paper.  The Egyptian hieroglyphs literally read mry ḥw.t-ḥr mfk3.t ("Beloved-of-Hathor [of?] Turquoise").  The Semitic alphabetic inscription reads (on both sides of the sphinx):

ḏ 'h[bb]clt
z nfk l bclt

The doubling of the b's is indicated both by context and possibly by a very unusual, possibly somewhat unique dot in the middle of the 'b'.  A similar dot occurs within the second c and both may suggest essentially consonantal reduplication.  This is only plausible if the first line refers to a very common epithet (which it does), in which the construct state had lost its inflection, probably due to common (perhaps cross-dialect) usage.  The epithet/title is thus in the simple nominative rather than in the genitive (through a construct state).

Then, similarly, the second occurrence of 'the Mistress' must actually be treated as a D-stem verb meaning 'to wed'.  The inscription is thus sort of bilingual, with the Egyptian actually being perhaps more ambiguous than the Semitic, and with the Semitic reflecting a mythological union between Set (most likely), The-One-of-Turquoise and The-Beloved-of-The-Mistress, and The-Mistress herself.

Sinai 345 also appears to be unique in using an 'Egyptian viper' (f ) to render this phoneme (/f/).

ḏu 'ahu[bb]acltu 
za nafki la baccilat

"This is The-Beloved-of-the-Mistress.
The-One-of-Turquoise does she wed."

Thematically, this inscription has at least one parallel.  This is from Note 199 in the larger paper: "Sinai 350 – R’L . ṢṮ . [Ẓ?]B[Ṭ?] . NQB / L’HB . [---]T – “The god Set is the Keeper? of Naqab / The-Beloved-of–[the-Mistr]ess.”"  It appears to reflect the sexual relationship perceived between Set and Zat Baalti, in the latter's aspect as a 'Mistress.'  The relationship with turquoise is unclear, but is apparently associated with Set.  I note elsewhere in the larger paper that Set's name is surprisingly similar to a Semitic root found basically in Akkadian (as well as Hebrew) meaning 'to shine.'

Sinai 345 also appears to evidence a monogrammed signature suggesting it was written by a scribe named Bin Surr, or more likely originates from a 'tribe of Surr,' quite possibly related to the city of Tyre.  A 'tribe of Surr' (most likely, i.e. BN ṢR or Bani Ṣur), occurs in several inscriptions.

[i]  John Coleman Darnell, F. W. Dobbs-Allsopp, Marilyn J. Lundberg, P. Kyle McCarter, Bruce Zuckerman, and Colleen Manassa, “Two Early Alphabetic Inscriptions from the Wadi el-Ḥôl: New Evidence for the Origin of the Alphabet from the Western Desert of Egypt,” AASOR 59 (2005): 90, accessed October 4, 2011, http://www.jstor.org/stable/3768583
[ii]  Primarily, Darnell et al. 2005; Gordon J. Hamilton, The Origins of the West Semitic Alphabet in Egyptian Scripts, The Catholic Biblical Quarterly Monograph Series 40 (Washington, DC, 2006).
[iii]  See also, Hamilton 2006, 327.
[iv]  Norbert Nebes and Peter Stein, “Chapter 7: Ancient South Arabian,” in The Languages of Syria-Palestine and Arabia, ed. Roger D Woodard (Cambridge, MA, 2008), 154.
[v]  See, Robert M. Whiting, Jr. “The Dual Personal Pronouns in Akkadian.” JNES 31, no. 4 (1972), accessed May 2, 2012, http://www.jstor.org/stable/543790.  The presence of this feature may make this one of the most archaic attested Semitic dialects ever recorded, certainly the most archaic West Semitic dialect.
[vi]  Herbert Bardwell Huffmon.  Amorite personal names in the Mari texts (Baltimore, MD, 1965), 107.
[vii]  CAD, I, 271-4, accessed April 30, 2012, http://oi.uchicago.edu/pdf/cad_i.pdf
[viii]  John Huehnergard, “Appendix 1: Afro-Asiatic.” in The Languages of Syria-Palestine and Arabia, ed. Roger D. Woodard (Cambridge, MA, 2008), 243.
[ix]  Dennis Pardee, “Chapter 1: Ugaritic,” in The Languages of Syria-Palestine and Arabia, ed. Roger D. Woodard (Cambridge, MA, 2008), 26.                                                                          
[x]  CAD, Q, 147, accessed April 30, 2012, http://oi.uchicago.edu/pdf/cad_q.pdf                                   
[xi]  Aïcha Rahmouni, Divine Epithets in the Ugaritic Alphabetic Texts. Handbook of Oriental Studies. Section 1, The Near and Middle East. tr. J.N. Ford (Leiden, 2007), 110, accessed October 4, 2011, http://books.google.com/books?id=j0SxrpjbwiEC&lpg=PP1&dq=divine%20epithets%20in%20ugaritic&pg=PP1#v=onepage&q&f=false
[xii]  ibid, 111.
[xiii]  A. Murtonen, Volume 3, Sections Bb. Root System: Comparative Material and Discussions. Sections C, D and E: Numerals under 100, Pronouns and Particles, Hebrew Material, vol. 3 of Hebrew in its West Semitic Setting. A Comparative Survey of Non-Masoretic Hebrew Dialects and Traditions. Part 1. A Comparative Lexicon (Leiden, 1990), 227.
[xiv]  William C. Hayes, “The Tomb of Nefer-khēwet and His Family.” MMAB 30, no. 11 (1935): 34, accessed October 4, 2011, http://www.jstor.org/stable/3255037
[xv]  D. H. Gordon, “Scimitars, Sabres and alchions.” Man 58 (1958): 24, accessed October 4, 2011, http://www.jstor.org/stable/2795551
[xvi]  Darnell et al. 2005, 85.
[xvii]  Some authors of Darnell et al. (2005) repeatedly stressed to me that the inscriptions were far apart.  However, Plate II (Darnell et al. 2005: 116) suggests they are at most two or three meters apart.
[xviii]  And for first causing me to seriously consider this possibility, I must thank Brian Colless.
[xix]  Could the suggested connection (below) to nocturnal drunken rituals suggest these were carved at night by torchlight?
[xx]  Hamilton 2006, 325.
[xxi]  CAD, R, 26, accessed April 30, 2012, http://oi.uchicago.edu/pdf/cad_r.pdf.
[xxii]  Cyrus H. Gordon, Ugaritic Textbook, Revised Reprint 1998, Volume 1 Part 2, Analetica Orientalia 38 (Rome, 1965), 75.
[xxiii]  Murtonen 1990, 151.
[xxiv]  John Huehnergard, “Review: A Grammar of Amarna Canaanite.” BASOR 310 (1998): 66, accessed October 4, 2011, http://www.jstor.org/stable/1357578
[xxv]  Michael C. A. Macdonald, “Chapter 6: Ancient North Arabian,” in The Languages of Syria-Palestine and Arabia, ed. Roger D Woodard (Cambridge, MA, 2008), 199.
[xxvi]  Nebes and Stein 2008, 162.
[xxvii]  Pardee 2008, 16.
[xxviii]  See, John C. Darnell, “Hathor Returns to Medamûd.” SAK 22 (1995): 54, accessed October 4, 2011, http://www.jstor.org/stable/25152711
[xxix]  John C. Darnell, “Volume 1: Gebel Tjauti Rock Inscriptions 1-45 and Wadi el-Ḥôl Rock Inscriptions 1-45.” UCIOP 119 (Chicago, 2002), 133.
[xxx]  John Coleman Darnell, “The Apotropaic Goddess in the Eye.” SAK 24 (1997): 47, accessed October 4, 2011, http://www.jstor.org/stable/25152728
[xxxi]  Hamilton 2006, 325.
[xxxii]  Huffman 1965, 240.
[xxxiii]  Interchange of Semitic ġ and Egyptian may not be unknown either (Wolf Leslau, “Semitic and Egyptian Comparisons.” JNES 21, no. 1 (1962): 48, accessed October 4, 2011, http://www.jstor.org/stable/543552).
[xxxiv]  Peggy L. Day, “Anat: Ugarit’s ‘Mistress of Animals.’” JNES 51, no. 3 (1992): 181-190, accessed October 4, 2011, http://www.jstor.org/stable/545543
[xxxv]  Darnell 1995: 48.
[xxxvi]  CAD, Ṣ, 123.
[xxxvii]  In other words the borrowed verb’s structure would parallel Akkadian ḫu ‘rest’ > nuḫḫu ‘appease’ (CAD, N, 143-50, accessed May 3, 2012, http://oi.uchicago.edu/pdf/cad_n1.pdf.).  While tempting to simply identify this verb with the Akkadian, the Semitic orthography is wrong (Sem. nwḫ vs. nwḥ).
[xxxviii]  Darnell 2002, 133.                    
[xxxix]  Hamilton 2006, 325.
[xl]  A. Jamme and W. F. Albright, “An Archaic South-Arabian Inscription in Vertical Columns,” BASOR 137 (1955): 34, accessed October 4, 2011, http://www.jstor.org/stable/1355555 .  But based on their identifications (in Jamme 863) of the composite letters of the monogram, this hypothesis makes little sense.
[xli]  Hamilton 2006, 328.
[xlii]  Eric L. Altschuler, “Gloss of One of the Wadi el-Ḥôl Inscriptions,” Ancient Near Eastern Studies 39 (2002).
[xliii]  See, Darnell 2002, 195-204.
[xliv]  ibid, 129-38.
[xlv]  Darnell 1995: 47.
[xlvi]  Hamilton 2006, 391.