CRYPTCRACKER: JERUSALEM JAR INSCRIPTION (2)

This connects with my own practice of drinking water with added cider vinegar (to kill germs and receive other benefits).

From her excavations at the Temple Mount (more particularly the Ophel) near the southern wall, Dr Eilat Mazar has brought to light two pieces of a pithos (a neckless ceramic jar) bearing a short inscription. The artefact is or was described here (audio-visually) by Eilat Mazar and Shmuel Ahituv.
    At one point Eilat reaches for her grandfather's book, and I was able to take my own treasured copy of it from the shelf right next to me, and turn the pages with her (Benjamin Mazar, The Early Biblical Period, 1986; King David's scribe, 126-138).
    As usual a priority claim is made: it is the "earliest" West Semitic (or even Hebrew) inscription from Jerusalem (though there is cautious hinting that it might be in an unknown Jebusite language).
    Actually, there is an older graffito from the Temple Mount that appears to belong to the Bronze Age. One suggestion I made for it was: YSh S.NR P, "There is a water pipe here", including the word used by David in 2 Sam 5:8, "water shaft". But it might be from the Iron Age, and the central three letters could be DWD (David).
    I was pleased to hear Shmuel Ahituv pronounce "Canaanite" (with the stress in the middle instead of the crazy English-speakers' practice of hitting the first syllable), and glad to see him personally pointing out with a stylus what he thought to be the direction of writing (left to right) and giving the identity of each letter:
   M, Q or R, P, H.(Het), N, break, L?, N (a reversed form of the previous N).
    (The R and Q on the Qeiyafa ostracon are also confusing to modern interpreters.)
    Allegedly, this combination of letters has no meaning in known West-Semitic languages.
    We'll see about that.
George Athas examined the picture and gave out his customary admonition : if you haven't seen the inscription itself, having only looked at photographs, you haven't seen the inscription. He proposed:
n lmnḥṣrm (note his Sadey for the P/L, which is not probable).
   But the second high-resolution photograph sent out by Dov Smith allows for detailed scrutiny.
   George Athas tried reading it from right to left, but Shmuel Ahituv says that the writing runs from left to right, opposite to the Hebrew Bible and Hebrew newspapers.
   Dextrograde (left to right) seems to have been the prevalent direction at that time.
    The Izbet Sartah ostracon has its abagadary and text dextrograde (sinistrodextral).
    The lines of the Qeiyafa ostracon (the David and Goliath inscription from Sha`arayim) each start on the left side.

    Here is my first view of the Jerusalem inscription (before I saw the other readings that have been issued, which are discussed below):
M R P H. N [N [L?] N
The first sequence could be divided into three words:
    ... M RP H.NN
"NICE PURE WATER"
M logogram or mu or [M]M 'water'
RP (root rp' "heal') 'pure, purified' (cp 2 Kg 2:21-22, Ezk 47:8-9, 'healed water')
H.NN (root h.nn 'favour') 'nice, favourable'
(LN could be 'for us')

   The characters seem to have been incised on soft clay, before the pot was fired. If so, the purpose of the vessel must have been known in advance.
   Incidentally, a Halif jar handle has on it (incised before firing) L N S. T ("for firing") (Colless, 1991, 50-51).
   Here Colless's cardinal rule of epigraphy must be reiterated: Only the person who wrote an ancient inscription knew what it means.
   And their shorthand system made it hard for us to get their meaning: there is usually no separation of words; there are no indications of vowels, only consonants; and this text is (typically) damaged, so we do not know whether the inscription is complete (there is certainly one vital piece missing from the text we have).
    However, my proposed reading (NICE PURE WATER, subtext: this pot is for clean water only; do not put milk or flour or flesh or fish in it) is almost self-authenticating. (!)
    This makes a change from LMLK jars (for or belonging to the king). Shmuel Ahituv said that the Jerusalem inscription could refer to the owner, or else the contents.
    Examples of such labels on jars are found in the Bronze Age. From Gezer there are storage vessels with early alphabetic signs on them, engraved before baking: two of them (9 and 10) have a single M, and another has two side by side (11-12) (Sass 1988, Fig. 248; Colless 1991, 20, 22; ); this might mean they were water pots. More on this below.
    There is a another early "Canaanian" inscription on a potsherd, from Beth-Shemesh, which has the word H.NN meaning gracious, but with reference to a voice, though there is a liquid connection; the voice is lubricated with wine (apparently YN, and not YYN, the Biblical Hebrew form).
   I have plenty of other suggestions in the bizarre category, for which I am famously infamous:
"Myrrh/bitterness for John" (H.nn) (taking the P as L, like the one on the Wadi el-Hol graffito, and the Aramaic Fekheriye inscription ). On the Hol inscription from the Egyptian desert, we know it is L, because P appears as a full mouth on the horizontal line [ (|) ].

Please note that my approach to Early Iron Age inscriptions is different from the method of other interpreters: they are working back from the first millennium forms of letters, while I am looking at the signs in their evolution from their pictorial beginnings in the second millennium BCE (see my table of signs below). Thus, P was originally (and obviously) a human mouth, and the lower lip was gradually lost, as perhaps we can see happening in the example here (perhaps P, possibly L, a herdsman's crook). My account of this development, with a chart showing all the letters, is available here.

[16th of July 2013 onwards]
   I am now in a position to respond to other suggestions that have been offered.
    Douglas Petrovich has collected various interpretations of the letters (Ahituv, Rollston, Demsky, Galil, himself, myself) and has placed them at our disposal here. [PS. This has been expanded and published in permanent print in PEQ 147, 2, 2015; my response is recorded below.]
   Looking back to get our bearings, here is an extract from the original announcement (sent by Dov Smith on 10th of July, and widely disseminated):
"The inscription is engraved on a large pithos, a neckless ceramic jar found with six others at the Ophel excavation site....The inscription was engraved near the edge of the jar before it was fired, and only a fragment of it has been found, along with fragments of six large jars of the same type. The fragments were used to stabilize the earth fill under the second floor of the building they were discovered in, which dates to the Early Iron IIA period (10th century BCE). An analysis of the jars’ clay composition indicates that they are all of a similar make, and probably originate in the central hill country near Jerusalem."
     If the inscription was incised into the clay before baking, then the text must have the same age as the pot itself.
    And we may assume that the object must be older than the building under which it was discovered?

    Here is my response to Christopher Rollston, who also attempted to read it on the 11th of July, and posted it on his site as The Decipherment of the New ‘Incised Jerusalem Pithos’.
            M Q L H. N [R [Sh
   For the five complete letters, I think his reading is possible.
    QLH. is understood as "pot", and NR as the name of its owner.
    If this pot is a pithos (a large storage jar, and note that in its Greek setting a pithos was usually for wine), what would be the Classical Hebrew word(s) for such an object?
    In any case, if this is not a cooking pot, then Christopher Rollston's QLH. should not work, as the meanings it has in its Hebrew history are "cauldron, kettle, pot for cooking".
    Also, whether in its original Egyptian setting, or in Coptic, or Ugaritic, or Hebrew, it has a final -t, and it is difficult to spot a Taw near the Het, except by deconstructing the Het and making its bottom right corner a cross (+), thus constituting a unique ligature (not impossible, I suppose).
    Incidentally, that is a strange Het, with two horns and two legs, and only two crossbars (as I argue, it started as a H.asir, a mansion with a courtyard), and we need to resort to the 9th-C. Moabite stone of King Mesha` to find a peculiar counterpart with the three characteristics we encounter here; but Ahituv shows two other similar examples (Batash and Eshtemoa`) on his comparative table (viewable here).
    I tentatively preferred R to Q, and P to L, producing:
     ... M R P H. N [N] [ ] N ...
     "Nice (h.nn) pure (rp) water (M logogram, or mu, or [M]M) ...."
    Contra Chris Rollston's M Q L H. N [R [Sh
    He goes to the Fekheriye Aramaic inscription for his Lamed, which is quite abnormal for its time; elsewhere, so as to distinguish them, P is upright, while the hook of the original crook of Lamed is at the bottom. That is the situation on the Qeiyafa ostracon, though there we find more than one form for P and for L.
    Rollston argues that there is indeed a Resh that can be distinguished from his Q, but it is a head with a large cleft in its top (a bit wider than on his drawing), and an unusually long neck. Actually, with these features it should be Waw, if only one letter is constructed from the remnants.
    In general, Q (a cord wound on a stick) is round at the top, whereas R (a human head) is angular; but the stem of Q sometimes moves into the circle.
However, supposing it is W, not R, and the last letter is N, not Sh, and with a Yod between them, we have WYN "wine"(attested as WN on the Wadi el-Hol inscription of the Middle Bronze Age; but in the Iron Age the West Semitic form was YYN (yayin), or YN (yayn) as on the Beth-Shemesh ostracon.
    Gershon Galil has a view of it which chimes in with his interpretation of the Qeiyafa ostracon (which he interpreted as a set of instructions for social compassion, regarding oppression of widows, orphans, aliens, slaves, and poor people). Reading from right to left:

[nt]n [tt]n ḥlqm Give them their share

The reference would be to "poor brothers".

But he issued another conjecture (also recorded with more detail here), again from right to left:

[…], mem, qop, lamed, ḥet, nun, [yo]d, [yo]d nun

... N YYN H.LQ M … spoiled wine from…

Douglas Petrovich has taken up this idea here, together with analogies from Egyptian practice in labeling jars of wine, adduced by Gershon Galil, and with speculation about additional words at either end. The text would have stipulated a regnal year of a king, and declared that this "smooth wine" (yyn h.lq) was "from" (m) a particular place.

   A difficulty with YYN (yayin) is (as noted above) that it has an additional Y, whereas Ugaritic texts and the Beth-Shemesh ostracon (where we see bt yn for byt yyn, "wine house") have simply YN, and so the form yyn suggests written Hebrew that is later than we would expect in the 10th Century BCE. However, I have proposed to read BYT in two fragmentary LBA inscriptions from Lakish (Colless 1991, 40-41). It is a matter of diphthong contraction (or whatever) rather than use of matres lectionis ; the second Y is pronounced; it is not there to indicate î.
    The two hypothetical Yods are very large on Gershon Galil's drawing (perhaps on account of the stray line poking out from the gap and reaching lower than any of the other letters, which might belong to another letter or could be a mere scratch). The letter Yod began as a hand with its forearm (yad), and in the Iron Age it looks like a reversed F; the YYN conjecture has them both as an inverted form of this; but the Qeiyafa ostracon has both types, as Galil would know.
   (Finally [January 2014], Gershon Galil has put his conclusions into print:
‘yyn ḫlq’ The Oldest Hebrew Inscription from Jerusalem, Strata 31 (2013) 11-26.
This seductive solution is discussed below, 25th of January 2014.)
     Another immediate attempt to reconstruct and transcribe the inscription was made by Reinhard Lehmann and Anna Elise Zernecke in KUSATU. The missing letters in the gap are possibly M and Sadey, hence:
     M Q P H. N M S. N   (or N S. M N H. P Q M)
The sequence S.N looks promising, as the name Siyyon (Zion) preceded by mi(n) "from"; so this would be the source of the contents of the pot?
    Without the missing piece, the right-hand end of the text is a mystery. Can it be found, please?
    After our experience with the Tel Dan inscription, are we certain the join has been made correctly? It looks like a good connection to me.
    But there is no indication of word-separation, and so there are too many variables (including the gaps) for us to be reading this text completely, or to be dating it to its precise decade.
    We cannot tell (unless the missing pieces turn up) whether there were other letters preceding and following the part that we have.
   Thus Aaron Demsky supposes a Het on the left, and produces the word h.mr "fermenting wine"("for Hanan").
    But I think that the proverbial COOL CLEAR WATER is a good candidate for the contents of the vessel, and that this was stated on it when it was first made; it was not for milk or meat or fish.
   It needs to be added (as already intimated above) that I got this idea from a set of Bronze Age storage jars from Gezer, which were likewise inscribed before being baked (Colless 1991, 22, depictions; 31, discussion). Some have M (a vertical wavy water sign), others have MM; presumably, in both cases "water" is meant.
   Other inscriptions in this collection, all single letters, may be abbreviations of the commodities they contained:
   Y (yn wine) H. (h.mr fermenting wine) T (trsh new wine) Sh (shkr beer or shmn oil) H (hlb milk) S (smk fish).
There are some other Bronze-Age vessels (of the pithos category) found in a cemetery near Tel Aviv (Colless 1991, 24, depictions; 51-52, discussion). One pithos has a vertical wavy line between a pair of strokes, possibly M, indicating that it was a water jar. Another has a long horizontal wavy line at the top, with a symbol below it: a cross inside a circle. This is an old form of Tet and Theta, and I presume it is a development of the Egyptian nfr sign (+o) which is attested in proto-alphabetic texts; it was used acrophonically for T. and logographically for T.ABU ("good and beautiful"); hence the Mem and Tet would say "good water".

    However, the contents of this rediscovered fragmentary Jerusalem jar are still to be ascertained.
    Nevertheless, there is another solution (bubbling up from a teeming source in my seething brain): if we accept the Sadey proposed by Lehmann and Zernecke, there are other possibilities besides "from Sion".
           M S. N [R] "from the spout"
   As I said above, an older inscription from that area of Jerusalem may have mentioned the "water-pipe" (s.nr), which features in the capture of Jerusalem by David (2 Sam 5:8); on this interpretation, water would be collected there in or for this pot.
     Or, invoking the root S.NN, denoting coldness (with MS.N as a participle, or S.N or S.NN as an adjective or noun, and supposing H. N M S. N or H. N N S. N) we have a very seductive combination of hypothetical words:
   M RP H.N[N] [S.]N[N]
             NICE (h.nn) COOL (s.nn) CLEAR (rp) WATER (M)

   In support of this reading, I could make the following case.
   There are many analogies for pots with the letter M for water marked on them (but there are examples of inscribed storage jars with indications of their contents, besides water).
   The direction of writing is not certain, but the general consensus is that it is dextrograde, though Galil and Petrovich are convinced it is sinistrograde, and they produce a tempting reconstruction of the text. For syllabic writing in the Bronze Age and consonantal writing in the Iron Age, the orientation style in Byblos (and elsewhere in the north of Canaan) was right to left (sinistrograde). On the other hand, in the south (including Sinai and Egypt) it is hard to see a consistent direction, but the few substantial documents we have from southern Israel in the early Iron Age, namely the Qubur el-Walaydah bowl, the Izbet Sartah ostracon, and the Qeiyafa ostracon, run their lines of writing from left to right (dextrograde), but the Beth Shemesh ostracon has five columns (right to left with the last one running back horizontally and to the right. Then the Gezer calendar appears, and it is conforming to the Phoenician style, not only in line orientation but in the forms of the letters, and this becomes standardized in Israel.
    This adoption of the Phoenician form of the alphabetic script could have arisen from the diplomatic and commercial relations David had with Phoenicia (2 Samuel 5:11-12), and likewise Solomon (1 Kings 5), which did not happen in the time of King Saul; that is my working hypothesis, and I hope it works.
   The question is: on which side of the dividing wall is the Jerusalem pithos inscription standing? We would like to know whether it is an Israelite inscription like the two five-line ostraca (Izbet Sartah, Qeiyafa); but even if it is Canaanian or Jebusian, we would still like to establish its position on the spectrum, to decide whether it is before King David's capturing of Jerusalem or after; more particularly, whether it was written before the official changeover to the standard Phoenician alphabet for writing Hebrew, the language of Israel. We have a set of clues to assist us.
   Here are some of the differences between the characters on the three ostraca versus the calendar (as a representative of the newly adopted Phoenician style, observable to some extent on the table appended below):
   `ayin is everywhere a circle in the Iron Age, though it had a more natural eye-shape in the Bronze Age; it has a distinguishing dot in its circle on the three ostraca, but this point is lacking in the calendar, as in the Phoenician inscriptions;
'alep will always have the original ox-head resting on its side with the snout pointing to the left (Qeiyafa has this, but pointing to the right, and also the original ox-head, and even the A that will turn up later as Alpha); the snout of the ox also indicates the direction of the writing, as does the triangular human head of the Rosh, and the two and three prongs of the Yod and He respectively;
Shin now consistently has the shape W or VV rather than 3 (Qeiyafa has both in the second line);
    Lamed will not have the crook at the top (Beth Shemesh, Izbet Sartah, though not Qeiyafa) but will constantly be L or l;
Pe is a mouth that has lost part of one of its lips; it is consistently curved or angled at the top, and this should settle the question whether the Ophel letter is L or P; the early southern examples of L, whether the crook is at the top or at the bottom of the stem, generally have the round part curling right in (Walaydah, Beth Shemesh, Izbet Sartah, and Qeiyafa in most cases); the numerous instances of the Aramaic L in the long Fekheriye text also curl inwards, and do not really support reading the Ophel letter as L rather than P; but P occurs so rarely that there are few instances for comparison, and it has to be said that no known P bends round as much as this character; still, the next two letters, certainly identifiable as H. (Het) and N (Nun), are not conventional, as noted above;
    Qop was originally qaw, a cord wound around a stick, with the stick protruding at the top; in the Iron Age it was reduced to a circle atop a stem, and on the Gezer text the stem has penetrated right into the circle, which is a Phoenician feature;
Rosh is indubitably a human head, and it is more triangular than circular (Q); but with only one of them present, R and Q are hard to separate in this case;
    Samek has a different letter (and here I am pointing out a significant detail that has been overlooked in the past): instead of the fish for S (as in the Sinai inscriptions, Beth Shemesh, and Izbet Sartah) the spinal column (Egyptian djed) becomes the standard Samek (the Qeiyafa ostracon apparently does not have this Samek, but I think the fish is in line 4);
Mem will always be vertical (falling water) not horizontal (level water).
    The M in the Ophel text is neither vertical nor horizontal, but is leaning to the right, as also the presumed R (the head is facing rightwards, and yet all the heads in other inscriptions, except the one on the Beth Shemesh ostracon, are looking to the left). The first snake (actually a set of water-waves) is apparently looking where the writing is heading, even though this position is anomalously unique for the Iron Age. But the human head and the snake are perchance pointing in the right direction.
Lacking 'alep, `ayin, shin, and especially samek, and with the Q/R and L/P confusion, as also the uncertainty over its traveling direction, this inscription is well and truly concealing its identity from us.
    For assistance in the inquest, we could call as witness the Kefar Veradim bowl (Rollston 27-33, with drawings of its inscription, and also the Gezer calendar and the Tel Zayit abgadary, and a discussion of their interrelationships). The bowl comes from a burial cave in northern Israel (between Akko and Hazor), and would date from the 10th C BCE (Iron Age IIA-B). Its inscription runs round in a circle, so it is hard to say whether its M is vertical or horizontal; it has the Phoenician features: Samek as spinal column, not fish; Shin as W not 3; K with no stem; but, to confuse the issue, its Het has no legs or horns, and surprisingly its `ayin has a central dot.

16th of November 2013
   Note that the final presumed N on the Ophel pithos inscription is a reversal of the previous one, and this could help us date the text. As I see it, there was a change in script in the time of David and/or Solomon, with the Hebrew version of the alphabet now conforming to the international Phoenician alphabet, which was a consistent consonantary with only one form for each letter, whereas in Israel in the era of the Judges and King Saul an inscription could contain two or three versions of a particular letter; for example, the Beth-Shemesh ostracon has 'ayin as a circle with and without a dot in it, and two forms of B; similarly the Qeiyafa ostracon has three different stances for 'A (also D, W, Y, L, T). We know that the consonantal alphabet of the 1st millennium BCE did not indicate vowels; but it now appears possible that Hebrew inscriptions in early Israel (Iron Age I) used the alphabet as a syllabary, with forms for ba, bi, bu, and so on. Hence, with two versions of N (or else M), this inscription might belong to the time before King David occupied Jerusalem. But it depends on the dating of the pithos, if that is possible.

25th of January 2014
As noted above, Gershon Galil has put his latest ideas into print:
‘yyn ḫlq’ The Oldest Hebrew Inscription from Jerusalem, Strata 31 (2013) 11-26.
The wine (yyn) is no longer "spoiled" (or "smooth" as Douglas Petrovich had it) but "low-grade". A counterpart for Hebrew YYN H.LQ is found in Ugaritic YN HLQ, which could mean wine that is "lost" or "perished" or "gone sour", and not fit for the king's table, but good enough for the workers and the soldiers. As he did with his interpretation of the Qeiyafa Ostracon, Gershon adds a wealth of background material, which will remain useful, even if his hypothetical reconstructions of the inscriptions turn out to be incorrect.
   Let us consider his summing up (p 22).
"The inscription also indicates that there were scribes able to write texts in Jerusalem as early as the second half of the 10th century BCE. These scribes may have been Canaanites or other non-Israelites since David and Solomon employed non-Israelites in their government, even in very senior positions, including the office of the Chief Scribe. David’s scribe was Shisha [or Shawsha or Shewa, perhaps a Hurrian name or title, and his Hebrew name Seraiah, BEC], and his sons were appointed as Solomon’s chief scribes (Mazar 1986: 111–138) [read 126-138, BEC]. So it would not be surprising that this inscription was written in the ‘Late Canaanite script’ (as was the Qeiyafa inscription), and that it indeed reflects the southern Hebrew dialect; but it also uses archaic technical terms like ḫlq to define this inferior wine.
"The forms and stances of the letters in this inscription (as well as in the Qeiyafa inscription) are not yet fixed – a phenomenon typical of pictographic scripts. But this inscription is written from right to left, while the Qeiyafa inscription runs from left to right. This fact may indicate that it is the beginning of the regulation of the reading direction."
GG sets the pot and its inscription in "the second half of the 10th century BCE", because this is a Type B pithos, "dated by archaeologists not earlier than the late 10th–9th century BCE" (p 13). May we ask whether this is high or low chronology? It is supposed to be Iron Age IIA (starting around 1000). Can we get clarification on this? How can it ever be claimed that this is a Jebus inscription, if the pot dates from the time of Solomon?
   GG speaks of "the southern Hebrew dialect" and insists that this always has YYN for wine, not simply YN, overlooking the Beth-Shemesh Ostracon, which apparently has BT YN for "tavern (wine house)" (Colless 1991, 45-49).
"The similarity between this [Qeiyafa] inscription and the new Ophel inscription is already well attested. Therefore, the following discussion will focus on the letter yod." (p 16)
   What similarity is meant? The two Yods on his drawing are hypothetical; the Nun, Het, and Lamed have no counterparts on his drawing of the Qeiyafa text (UF 41, p 196); and Mem is always a zigzag. No matching forms are observable on Phoenician inscriptions, either; nor on Hebrew inscriptions in Phoenician style, such as the Gezer Calendar and the Tel Zayit Stone. Perhaps this is the work of a scribe who had his own personal version of the alphabet, or else we are looking at forms that have not cropped up in extant inscriptions, but each has its place in a larger scheme, such as a syllabary with 66 signs, not merely the 22 letters we know from the Phoenician consonantal alphabet.
   GG says that the forms and stances of the letters "are not yet fixed", and he offers an analogy with "pictographic scripts"; true enough, but we are a long way from that stage of the alphabet (see the table below). As I have just suggested above, these variations in form and stance may not be arbitrary but significant: the alphabet was functioning as a syllabary in that period.
   I have already made detailed comparisons of the characters in the inscriptions from Iron Age I and IIA, and I have a tentative table in my work-in-progress sheets of paper. I would like to apply these syllabic values to the signs on this Jerusalem jar, as reconstructed by Gershon Galil. I do this in the spirit of Rabbinic discourse, where one scholar will give support to the arguments of his cobber (h.br), even though he himself is following a different line of reasoning, and, of course, I have argued here (but only hypothetically and tentatively, and without conviction) for a water-jar interpretation, rather than a wine-vessel approach.
   Reading from the right (but setting aside the M or N): YA YA NU H.U LU QU.
The two Yods (as plausibly reconstructed) are not really different, and can indeed be compared with one of the three Qeiyafa forms (the Yod at the end of line 2, as GG says) the one which I propose as YA. But if the word is YAYIN, then we would expect the second Yod to be different. However, in syllabic writing when confronted by a word like YAYN it is customary to give the vowelless syllable the same vowel as a neighbouring syllable, and the same will occur in the accompanying word H.ULQU. The -u vowel on the end of each word signifies the nominative case (or it is a "dead vowel", if case endings had ceased to be used by then).
   The presumed LU-sign is an inversion of the standard Phoenician L. The QU with a sloping stem is actually found near the end of Qeiyafa line 5, but GG has chosen to identify it as R there, and he wants the R (with an upright stem) in line 4 to be Q. The forms of the two Mems do not match perfectly with the standard models
    Accordingly, the whole line could read: -ma yaynu h.ulqu mi-
And it tells us ("formulaicly") that this jar was made in a particular regnal year of a particular king to contain vinegary wine from a specified place.
Regarding the direction of the writing, GG suggests "it is the beginning of the regulation of the reading direction", and so he apparently recognizes that the turnabout was the result of a decree. Nevertheless, the writing system in this inscription is still the old style, which I am inclined to call the Neo-syllabary (distinct from the original West Semitic syllabary, particularly connected with Byblos).
My watered down version does not work so well, if RP is actually LQ.
   MI QU LU H.U NU . . NI
    And forcing the root H.NN (be gracious) to apply to water is not natural or normal or nice (but "nice" is what I want it to say).

12th of July 2014. It has occurred to me that if Gershon Galil's proposed restoration of two yods is right, then we have another indication that this is before "standardization" (and thus neo-syllabic): that particular stance is not found in the international consonantary, but it appears on the Qeiyafa ostracon, and that is where Gershon got the idea. Also, given the possibility that the letter to the right of the YY could be N rather than M, a reading YYN is possible whichever direction the writing is running.

3rd of August 2014. A reason for the oddness of the letters of this inscription (when treated as syllabic) is that there are so many -u syllabograms: QU LU H.U NU, which differ from the a-forms of the Izbet Sartah abagadary, and the i-forms of the standard consonantary.

   Much ingenuity has been applied by scholars to this bunch of ancient letters on the Jerusalem jar, but we can not attain certainty, because of the many variables (not least of which are the many missing letters) since the writer (who by my first principle was the only person who knew the meaning) has not obeyed my fundamental rule, that every early alphabetic inscription should be accompanied by a copy of the complete alphabet in the scribe's own handwriting, so that we can tell which letter is which (as in the Izbet Sartah text, but even there we are still left with puzzles by its author, who actually says "I am learning the letters" [ ' LMD ' TT ]); and he should have shown us 44 more versions of the letters, if he was actually using the alphabet as a syllabary with 66 characters.
   However, this is what we have so far: an incomplete collection of scattered bones from which to construct a dinosaur. But, as I now see it, these are the clues to work with: the Izbet Sartah alphabet (abagadary) represents that scribe's set of -a syllable-signs; the Phoenician consonantal alphabet is using the -i signs to represent all the vowels (or vice versa); the remaining signs (including four on this Jerusalem jar) are the -u syllabograms. Those are the new rules for West Semitic epigraphy with regard to Iron Age I inscriptions. I have said more about this on the ASOR site, now being updated here.

15th of July 2015 As noted above (post-scriptually) Douglas Petrovich has published his study of the inscription: Palestine Exploration Quarterly 147, 2 (2015) 130-145.
   He is substantially supporting Gershon Galil's reconstruction and interpretation of the text (but "inferior wine" becomes "pseudo-wine", which might be going too far in downgrading the liquor):
(L < R) ...N YYN H.LQ M...
"[In the firs]t [(regnal) year]: pseudo-[wi]ne from [the garden of ??]"
This is set out on one of his clever icons (Fig. 4) with transcriptions in Hebrew characters.
For bashana ("in the year") he rightly omits the final -h of the Massoretes as anachronistic, but his inclusion of ha- (definite article, in ha-ri'shona, 'the first') is also premature for this period of the Hebrew language (it is not yet attested at that time).
   He assumes the hypothetical "first year" would apply to Solomon's reign.
   For a comparative table of signs he reproduces in Fig.2 the chart of Shmuel Ahituv from IEJ. This is not entirely helpful; as there are only five complete letters, and two or three incomplete signs, the six consonants offered are insufficient to distinguish R/Q/ and P/L, and to illustrate the two reconstructed Yods. It is gratifying to see my own table of the evolution of the alphabet presented there as Fig. 5, but it lacks the details needed for this exercise.
   He has opted for two forms of Nun, and this suggests to me that the text is syllabic.
   He confuses language and script on p. 139, with regard to the terms Phoenician and late-Canaanite (as applied to types of script, not language). He insists that the language is Hebrew, as also on the Qeiyafa ostracon (with its SH-P-T., `BD, and MLK, but Rollston has questioned this). However, it seems that all these new inscriptions are in the language of Canaan, which even Philistines (and presumably Jebusites) were using to communicate. And on p. 142 he emphasizes his point that the newly-discovered inscriptions "represent an earlier phase of the Hebrew language". But he includes as evidence the alphabets (abecedaries) on the Tel Zayit stone and the Izbet Sartah ostracon, which can not tell us anything about language or dialects. (Not all texts written in Roman letters are Latin language.)
Douglas admits (p. 141) that "several readings are possible", but his is "the best option". If I had to make a choice, I would put it first as the most plausible, but still hold that my own suggestions are not impossible. This is a nice defense of an attractive hypothesis (with a good discussion of the dating of the pithos), but the inscription, as we now have it, is illegible (incomplete letters) and unreadable (no correct interpretation is possible).

Once again, when a cluster of scholars can not agree on the interpretation of an ancient text, we seem to have glaring proof that epigraphy, despite its refined academic jargon and its cautious technical methodology, is not an exact science. This new inscription has made fools of us all. Of course, if we had the missing pieces of the pithos puzzle we might be able to do better, so we live in hope that they may be found in the searching and sifting.
Under the present circumstances it is absolutely impossible to read this incomplete inscription; but it is quite possible that my reading of it was on the wrong track (direction-wise and liquidly). No sour grapes, and no champagne, but my personal secret is that I drink my water with a dash of vinegar (apple not grape).

9th of January 2019
Well, here I am, still alive: I have not perished (hlq) in the meantime.
Raz Kletter has published some notes on the Ophel pithos inscription, in PEQ (2018) 265-270.
   My first observation is that he misspells my name (as Colles, instead of Colless) on p.269, in his reference to this cryptcracker essay; that double S marks our family out distinctively amidst the Coles, Collis, Colles variants; but he has it right on p.265.
   While I am grateful for the mention, he is able to dismiss my contribution to the discussion, by declaring web-published readings as 'preliminary' (fair enough, mine are certainly tentative, and to be considered as work in progress); and my suggestion ('nice, cool, clear water') "finds no parallel in the form of Iron Age wares from Palestine inscribed with a label 'water' ". This sounds suspiciously like an argument from silence, and yet I gave two instances (five inscriptions in total) from the Late Bronze Age: from Gezer (M and MM for 'water') and Tel Aviv (M for 'water', and M T. as logograms or abbreviations for 'good' and 'water'). Recently I had occasion to read out in church the story of the wedding at Cana in Galilee, involving water jars having their contents turned into wine (Gospel of John, 2:7-9).
[problem: singular forms of adjectives with plural noun mayim?]

This script is not the Phoenician (or international) consonantal alphabet of the Iron Age (as on the Gezer Calendar and the Zayit Stone, and the Phoenicia column of my table, below): its letters are anomalous.
   First, the Het disqualifies it from that category; its lack of a third cross-stroke in the middle corresponds to the Het on the Moabite Stone, but none of the other letters has a counterpart there.
   Second, the Nun next to the Het is the wrong way round, though the incomplete character on the far right could be the standard Nun, or else a Mem.
   Third, the stance of the letter on the left of the Het is the reverse of the standard form of Pe, or else it is an inverted Lamed.
   Fourth, the Mem on the far left is almost correct, but it seems to have six strokes instead of five.
   Fifth, the two hypothetical Yods are only found on the Qeiyafa Ostracon, as one of three variant forms.

Although its presumed direction of writing (sinistrograde, right to left) fits the style of the international consonantal alphabet, its characters are so unusual that it could well be neo-syllabic.

17 May 2019
My latest suggestion
Reading syllabically from right to left (sinistrograde):
... N [ ] [YA] NU H.U LU QU (U?) M ...
"... bad/sour (h.ulqu) wine (yanu) and (u?) water (M...)"
This interpretation follows the line laid down by Gershon Galil and Douglas Petrovich: in their view the pot was a vessel containing inferior wine, possibly for workers on building sites in Jerusalem.
   Perhaps the full text contained a word for "jar" or even "pithos"; Ugaritic dn (also Arabic and Aramaic) was a jar for wine (and vinegar, Yoma 28b, Jastrow 315a). But Raz Kletter (268) says "Galil's reading does not fit the Jerusalem pithos, since it is not a wine vessel", and Kletter maintains that position because it has "a very large, open mouth", and this would allow the wine to have contact with oxygen and be ruined. This might support my first idea that it was for storing water, rather than wine; but if this was intended to be a container for spoiled wine, which was placed on open access for drinkers to put a ladle or cup into it, then there can be no objection.
   Kletter dismisses my water-jar solution because I cannot adduce such an inscribed vessel from the Iron Age, although I can find apparent examples in the Late Bronze Age (from Gezer, see above).
   Similarly, he points out that the term h.lq is not found with reference to wine in any other inscription in Palestine/Israel, nor in the Bible. But this could be an isolated and welcome instance of that usage, already attested at Ugarit; and maybe the language of Yebus (Yerushalayim) was closer to Ugaritic than to Israelian Hebrew. At the same time he ridicules the thought that a container of wine could be labeled as "inferior" or "bad". But that would only apply if the commodity was for sale, not if it was being offered gratis. This jar was perhaps made and marked as a receptacle for such liquid, and so it would probably not have royal references on it.
   Precisely what was meant by yn hlq remains unclear. If the primary sense of hlq is 'perish', parallel with mt 'dead', and the opposite of t.b (Cyrus Gordon, Ugaritic Textbook, Glossary, 969, p. 403) then such wine (a 'perishable' food) would be 'perished', 'gone bad', or 'become rotten', no longer 'good' (sweet) and thus 'sour' (vin aigre, vinegar). Biblical Hebrew h.lq does not offer obvious assistance, expressing ideas of smoothness or apportioning; but Petrovich leans towards associating the term yyn h.lq with deception (as in smooth-tongued), and proposing the improbable 'pseudo-wine' as the solution. He does mention Ah.ituv's note on h.ms. (vinegar) in Arad Ostracon 2 (Echoes from the Past, 98): Roman legionaries drank posca, "a mixture of vinegar and water, sometimes sweetened with honey". This reminds me of my own daily beverage of water with apple cider vinegar and honey. Of course, wine was customarily drunk with water. I am pondering whether these three ingredients can be found here: water (which I had first suggested as the contents of this vessel) with vinegar, and perhaps also honey.
   The water could easily be found in the final sequence M[..], as Hebrew mayim, or simply M as a logogram, or an abbreviation. It is more difficult to make this a 'honey pot', finding a place for West Semitic dbs (debash) or nbt.
   But if we are to read the combination "vinegar and water", where is the conjunctive W (or P)?
(The conjunction wa often seems to be lacking: for example, in the Wadi el-Hol list of sacrificial foods for the `Anat celebration.) It has occurred to me, that the missing WA is actually U (which would be expected here before labial M, in Hebrew) and this could not be represented in consonantal writing; if 'u was written, it would mean 'or'.
   I still need to justify the syllabic reading [YA]NU H.ULUQU, as opposed (but not violently so!) to [YY]N H.LQ.
   For his YYN, Gershon Galil posits two cases of Yod in the style of a character that appears three times on the Qeiyafa ostracon (though Yod also has two other stances in that neo-syllabic text): the arm has the hand (two strokes) pointing downwards, and that would make a total of four vertical lines, one of which is visible, emerging from the gap. However, these reconstructed figures are huge in comparison with the Qeiyafa model; that is not impossible, and its plausibility could be tested if the missing piece of the pot turns up; but I am proposing the alternative prototype of Yod with the hand at the top (as apparently on the Izbet Sartah ostracon, and on the Lakish bowl sherd); the two strokes of the hand (looking like pincers) are partly discernible above the empty space, with the end of the arm protruding at the bottom. However, this leaves a space before the Yod, increasing our frustration that the missing pieces are not available for inspection. There could have been another Yod there, as Gershon Galil suggested, indicating the southern Hebrew form of the word yayin.
   Regarding the syllabic reading, most of the letters are not the standard forms of the international alphabet, and by my calculation there is a preponderance of -u syllabograms in [YA]NU H.ULUQU.
Cyrus Gordon (Glossary 402b) has an Akkadian form hulqu (equated with la t.âbu, 'not good'), and H.ULUQU is a correct syllabic transcription of that.
   One little defect remains to defeat us: there is a space before the incomplete YA, and it could be another YA, producing [YAYA]NU, yaynu, 'wine' (Judean style). To achieve [YAYI]NI, yayin, the text would have to be read in the opposite direction, but this would annihilate H.ULUQU.
   If the first letter on the right is N, rather than M, and if it is syllabic and not simply the standard Phoenician Nun, then it possibly represents NI, having the reverse form of the other N, which I take to be NU. Could it have been preceded by the syllabic sign HI? Possibilities are: HINI hin measure? hén? hinnè? "Here is" or "This is"?
   The letter He appears twice on the incomplete Beth-Shemesh vertical inscription, engraved on two shards from a vessel. The second is in a sequence HN, accepted by P. Kyle McCarter as the measure hin. The first H has the form of Greco-Roman E. The second H seems to be the same, though the photographs show a projection from the bottom of the spine; this would make it the same as the standard Phoenician form, though reversed (its three strokes could point in the direction of the writing, to the right or to the left, along a horizontal line).
   Kyle McCarter's drawing shows the projecting line as a surface defect in the clay (although it seems to be attached to the letter in photographs) (McCarter, 188, Figure 5, and 185, n.2). He favours a meaning hin (measure) for the word, and this syllabic sign would provide the HI syllable. When there are a few letters only, it is difficult to establish whether the text is syllabic or consonantal, and that is likewise the problem with the Ophel pot.
    From all the sifting of ancient debris of Jerusalem, the missing pieces of the puzzle might turn up, but we shall probably never know what the scribe intended. Meanwhile I am holding onto my guess that this was an open-mouthed jar for containing water mixed with vinegar, for the refreshment of troops of soldiers, and/or gangs of workers. If it is syllabic it is Iron I, like the Qeiyafa ostracon; if it is simply consonantal, but not conforming to the the standard international (Phoenician) alphabet, then it is to be classed with the Eshbaal jar from Khirbet Qeiyafa (Shaarayim), to be dated in late Iron I rather than Iron II, in the time of King Saul, not King David, nor King Solomon, and specifically in the Yebus period of Yerushalayim, before David annexed it to his united kingdom.
   For clarification, let me state my present position on Khirbet Qeiyafa and its two inscriptions: the text of the ostracon is neo-syllabic, and its subject is the victory of David (not King David) over the `Anaq named Guliyut; the Eshbaal jar inscription is consonantal, but its letters are not in conformity with the international alphabet used in the Levant, of which the characters tend to be the -i syllabograms in the neo-syllabary. The archaeological evidence from Khirbet Qeiyafa (Shaarayim) shows that its existence as a fortress in the Iron I period was of short duration; the presence of Eshbaal, a son of King Saul, on this site indicates that this town was built by King Saul, not King David, as a bastion against the Philistines of Gath and Ekron; but it was apparently destroyed by them, together with the capital city Gibeah (Tell el-Ful), at the time of the battle of Mount Gilboa (1 Samuel 31:1-7); less probably, it might have been a casualty in the war between the house of Saul and the house of David (2 Samuel 3:1). David chose not to rebuild Shaarayim, but moved further down the road to make Yerushalayim his stronghold.

   My continuing struggle with the Qeiyafa inscriptions is available for inspection:
https://sites.google.com/view/collesseum/qeiyafa-ostracon-2

27.6.2023
Daniel Vainstub has an interesting take on the Ophel jar: he reads the letters as Arabian, which is not impossible, and he finds a reference to incense from Saba.
https://www.academia.edu/99586646/Incense_from_Sheba_for_the_Jerusalem_Temple

References:
Benjamin Sass, The Genesis of the Alphabet (1988)
Brian E. Colless, The Proto-alphabetic inscriptions of Canaan, Abr-Nahrain (Ancient Near Eastern Studies) 29 (1991) 18-66.
Gershon Galil, ‘yyn ḫlq’ The Oldest Hebrew Inscription from Jerusalem, Strata 31 (2013) 11-26.
Douglas Petrovich, The Ophel Pithos Inscription: Its Dating, Language, Translation, and Script, Palestine Exploration Quarterly 147, 2 (2015) 130-145.
Raz Kletter, Notes on the Jerusalem Iron IIA pithos inscription, Palestine Exploration Quarterly 150, 4 (2018) 265-270.
P. Kyle McCarter et al, An Archaic Ba`l Inscription from Tel Beth-Shemesh .......

http://cryptcracker.blogspot.co.nz/
https://sites.google.com/site/collesseum/

Brian Colless
School of Humanities, Massey University, NZ

(Click on this table of the evolution of the alphabet to view it in enlargement)