By P.J. Williams, reproduced with the kind permission of E.J. Brill.






In 1 Kgs ix 17-18, we have a list of places that Solomon “built”. The passage reads, “And Solomon built Gezer and Lower Beth-horon and Baalath and tmr (kethib) in the steppe, in the land”. The qere at this point reads tadmōr (tdmr) as does the parallel passage in 2 Chron. viii 4. It is usually held that tmr is to be vocalized tāmār and is a location in southern Judah, perhaps Hazazon-tamar (2 Chron. xx 2) or [[263]] Tamar of Ezek. xlvii 19 and xlviii 28. In this view tāmār is an older form which the Chronicler altered to Tadmor (i.e. Palmyra) in the North as part of an aggrandisement of the extent of Solomon’s reign. The form tadmōr was then taken from Chronicles to be the qere of Kings. There are still some questions outstanding. The significance of the phrase bmdbr b’rs “in the steppe, in the land” has not been explained, nor has the connection felt by the Chronicler between tmr and tadmōr. A.B. Ehrlich is among many names who have suggested that a place-name is missing after b’rs.[1] But one does not need this suggestion to consider the phrase “in the steppe, in the land” as a relocation phrase, indicating that the place-name it follows is in quite a different location from those previously mentioned, i.e. not necessarily in Judah. Perhaps the definition of “the land” in Kings is extensive, and could include Palmyra (cf. 1 Kgs v 1 Heb.). We must also ask what the kethib is. If it is tāmār then the name means “date-palm” and may thus be connected with “Palmyra” which is also connected with “date-palm”. Whatever the origin of the name “Tadmor”, which is well attested in the second millennium B.C., it was later understood to be connected with the word “palm”. Julius Fuerst’s Lexicon shows that a connection was felt between Arabic tadmōr [sic, read tadmur] and “palm” since Palma in Majorca was translated tadmīr in Arabic.[2]

We here consider whether the Greek versions shed any light on the situation. Some Greek evidence supports the qere, such as Lucian θοδμορ and Josephus θαδάμοραν (Antiquities VIII 154). The Old Greek at 1 Kgs ii 24d has θερμαι, while at x 22a it has ιεθερμαθ. Similarly the Hexapla has θερμαθ. These forms are not to be explained as representing tdmr with a resh-for-daleth interchange. They are rather to be explained as representing trmd and trmwd, i.e. Tarmad/Tarmod which is the usual way in the Babylonian Talmud of referring to Tadmor/Palmyra. We should reject the idea of confusion of resh and daleth since, although it is certainly possible in the Old Greek, we would not expect it to occur twice in the Old Greek and to be perpetuated in the more careful Hexapla. In addition this view does not explain the theta at the end of ιεθερμαθ in Old Greek x 22a,[3] nor the theta at the end of θερμαθ in the Hexapla. This theta can, however, be explained as representing the final daleth of the form trmd. In addition the Hexaplaric ms. 247 has the form θερμωθ, which seems to be an attempt to bring the pronunciation of the word in line with the pronunciation of the final vowel of “Tarmod”.[4] Thus the Greek versions do three things. They attest the presence of the qere “Tadmor”, they identify the loca- [[264]] tion as Palmyra (Tarmod) and they testify to considerable fluidity in the name of the same location. None of them attests the kethib tāmār.

Looking at the manuscripts in Kennicott for 1 Kgs ix 18 we also see considerable fluidity.[5] While most have kethib tmr, fifteen have tdmr in their first hand, and a further six and two margins have the same word written plene: tdmwr. One manuscript and one margin have the form trmd, i.e. Tarmod or Tarmad and a further two have the same written plene: trmwd. One manuscript has tmd which may be the Arab tribe Thamod connected with Tarmod in B. Jebamoth 17a. Surprisingly, 5 manuscripts have in their first hand the form tmwr. The fact that five manuscripts record this reading, which is not an obvious scribal error, must be taken seriously. It seems to confirm the long-forgotten view of Heinrich Ewald which supposed that the kethib of 1 Kgs ix 18 was tammōr not tāmār.[6] In this form the daleth has anomalously been assimilated to the mem, perhaps to avoid three dentals following each other in the form ’t-tdmr. Under this hypothesis tammōr is simply a by-form of tadmōr. This explains a number of things better than the existing hypothesis that the kethib is tāmār. First, it explains the shift by the Chronicler or his Vorlage from tmr to tadmōr more easily. Secondly, it explains why we have no versional evidence of the pronounced kethib. If the kethib was simply a by-form of the qere and pronounced similarly we would not expect it to be transcribed into any of the versions. On the other hand if it was pronounced tāmār we might expect this to show up in some version. Thirdly, if the kethib is tammōr, we may explain not only the Hebrew forms in the Kennicott manuscripts, but also their relative frequencies. Of those manuscripts which wrote “Tadmor” as the kethib a fairly high proportion went further and wrote this in its plene form. This is because they were non-conservative manuscripts and so fluctuate considerably. On the other hand, of those manuscripts which continued to write the kethib tammōr, only a very small proportion, but still significant enough to attest the form, wrote it in its plene form tmwr. This is because they were generally conservative manuscripts. The five manuscripts which added the waw, may have done so accidentally because they knew what the form sounded like. The only thing that stands against this hypothesis is that the form tammōr is not well attested. However, those who do not accept this hypothesis must suggest an alternative explanation of the form tmwr in the Kennicott manuscripts.


[1]Randglossen zur hebräischen Bibel 7 (Leipzig, 1914), p. 235.

[2]A Hebrew and Chaldee Lexicon (4th edn, tr. S. Davidson; London, 1871), p. 1459.

[3]The ιε [at] the beginning of ιεθερμαθ probably represents the Hebrew object marker ’t.

[4]Interestingly the Syrohexapla has here tdmwr. This attests the interchangeability of some of these forms.

[5]Benjamin Kennicott, Vetus Testamentum Hebraicum cum Variis Lectionibus 1 (Oxford, 1776), p. 622.

[6]Biblische Jahrbuch 6 (1854), p. 89.

Note additional to published version: the suggested phonetic change from tadmōr to tammōr is in an almost identical phonetic environment to the attested change in Babylonian Jewish Aramaic, Jewish Palestinian Aramaic, Samaritan Aramaic and Christian Palestinian Aramaic from qVdm to qVm(m) in the preposition ‘before’, or the adjective ‘first’.