Professional Documents
Culture Documents
George Athas-The Tel Dan Inscription - A Reappraisal and A New Interpretation (Journal For The Study of The Old Testment Supplement) (2006)
George Athas-The Tel Dan Inscription - A Reappraisal and A New Interpretation (Journal For The Study of The Old Testment Supplement) (2006)
George Athas-The Tel Dan Inscription - A Reappraisal and A New Interpretation (Journal For The Study of The Old Testment Supplement) (2006)
SUPPLEMENT SERIES
360
Editors
David J.A. Clines
Philip R. Davies
Executive Editor
Andrew Mein
Editorial Board
Richard J. Coggins, Alan Cooper, J. Cheryl Exum, John Goldingay,
Robert P. Gordon, Norman K. Gottwald, John Jarick,
Andrew D.H. Mayes, Carol Meyers, Patrick D. Miller
12
General Editors
Thomas L. Thompson
Niels Peter Lemche
Associate Editors
Mogens Millier
Hakan Ulfgard
George Athas
ISBN 0-8264-6056-9
CONTENTS
Chapter 1
INTRODUCTION 1
Introductory Remarks 1
Methodology 1
Chapter 2
ARCHAEOLOGICAL CONTEXT OF THE FRAGMENTS 5
Introductory Remarks 5
Fragment A 6
Fragment Bl 13
Fragment B2 14
Synthesis of the Archaeological Data 15
Chapter 3
EPIGRAPHICAL ANALYSIS 18
Introductory Remarks 18
Fragment A: Physical Characteristics
and the State of Preservation 18
Fragment A: Epigraphical Analysis 23
Fragment A: Assessment 72
Fragment B: Physical Characteristics
and the State of Preservation 73
Fragment B: Epigraphical Analysis 78
Fragment B: Assessment 92
vi The Tel Dan Inscription
Chapter 4
PALAEOGRAPHICAL ANALYSIS 94
Introductory Remarks 94
Fragment A 96
FragmentB 137
Synthesis of the Script of Fragment A
and the Script of Fragment B 164
Summary of Script Analysis 165
Chapter 5
ARRANGEMENT OF THE FRAGMENTS 175
Evaluation of the Original Arrangement by Biran and Naveh 175
A New Arrangement 189
Chapter 6
TEXTUAL ANALYSIS 192
Introductory Remarks 192
The Text 193
Commentary on Fragment A 194
Commentary on Fragment B 230
Language of the Inscription 245
Grammatical Survey 246
Glossary of the Inscription 252
Chapter 7
HISTORICAL COMMENTARY 255
Introductory Remarks 255
Historical Considerations 255
Bayt-Dawid and the Quest for King David 298
Cultic Implications 309
Chapter 8
CONCLUDING REMARKS 316
Bibliography 320
Index of References 327
Index of Authors 329
LIST OF FIGURES
George Athas
Sydney, March 2002
ABBREVIATIONS
Grammatical Abbreviations
1+ +++++++++++
2. second person
3. third person
abs. absolute state
adj. adjective
coll. collective
com. common (gender)
conj. conjunction
cstr. construct state
del. deity
fem. feminine
gent. gentilic, of a people
impft. imperfect
inf. infinitive
juss. jussive
loc. location
masc. masculine
n. noun
nota ace. nota accusativa, object marker
pers. personal
pft. perfect
pfa+ +++++++
pi. plural
pr. proper
prep. preposition
pron. pronoun
sfx suffix
sg. singular
vb. verb
Chapter 1
INTRODUCTION
Introductory Remarks
The Tel Dan fragments are three broken pieces of basalt rock with carvings
of ancient Semitic writing on them. The first fragment was discovered in
1993 during excavations at the site of Tel Dan in northern Israel. The
second and third fragments were discovered separately in the following
year. In the few years since their discovery, these fragments have spawned
a veritable library of literature from scholars all over the world writing in
many different languages. The interest that these fragments have generated
demonstrates their value in the eyes of the scholarly community.
However, they have not gone without controversy. Indeed, such wide
and varying interpretations exist regarding the fragments that one can
easily get lost in the complexity of the issues. Most of the literature has
tended to concentrate on one or two principal issues regarding the frag-
ments. To date, no one has attempted a thorough and definitive analysis of
the fragments. I hope that this study will fulfil that need.
For the purposes of this study, I have chosen to follow the original
labels for the fragments, which the publishers, Biran and Naveh, first used.
This has been done in order to achieve a connection and easy reference
with the literature produced thus far. Therefore, the first and largest frag-
ment discovered is labelled Fragment A, the second is labelled Fragment
Bl, and the third is labelled Fragment B2.
Methodology+
The controversy that has ensued since the publication of the Tel Dan
fragments has highlighted the need to streamline a methodical approach to
studying the fragments. In 1995, Demsky proposed a five-step model in
2 The Tel Dan Inscription
special reference to the Tel Dan fragments.1 Demsky's article was mostly
written before the publication of Fragments B1 and B2 with only an adden-
dum making reference to these further discoveries. As such, his model was
produced with only Fragment A in mind. Nevertheless, Demsky aimed at
presenting a systematic model for studying all ancient Northwest Semitic
inscriptions. The five steps of his approach are:
1. To establish the archaeological context in which the inscription
was found.
2. To determine the type of inscription which is indicated by the
writing surface. This then allows the researcher to compare the
inscription with other known Northwest Semitic inscriptions of
the same genre.
3. A palaeographical analysis.
4. A linguistic analysis.
5. An historical synthesis.
While I agree with the general outline of Demsky's approach, I believe
there is room for improvement. Since the Tel Dan fragments have idiosyn-
cratic issues, such as what relationship each of the three fragments have to
each other, these have to be dealt with in an unbiased and systematic way
prior to final interpretation. The following is an outline of the approach
that this study will take:
1. To establish the archaeological context of each of the three
fragments discovered at Tel Dan. The purpose of this will be to
indicate whether the fragments were found in primary, secondary
or even tertiary usage. This involves examining the precise loca-
tion in which each fragment was found and determining a date
for the immediate context through the witness of pottery assem-
blages and structural remains. The ultimate aim of this first step
will be to provide a chronological bracket in which we may place
each fragment, if that is at all possible.
2. To conduct an epigraphical analysis in which the characters and
text on each fragment are itemized and clarified. An epigraphical
analysis will also serve to highlight how the text on each fragment
was produced. By comparing the results from each fragment,
I shall come to a preliminary conclusion on the relationship
It will become evident that Steps 6 and 7 will overlap significantly since
these two steps are effectively the goal of my research. Thus, the conclu-
sions from one of these steps will often be interspersed with those of the
other.
It cannot be stressed enough that this type of research demands an up-
close analysis of the actual fragments. Reliance on photographic images of
the fragments simply does not afford the researcher the necessary depth
for a comprehensive epigraphical analysis. Crucial aspects of a fragment
can be hopelessly confused, misportrayed or altogether missed by a two-
dimensional image in which only one configuration of light can be main-
tained. The same comments must be reserved for hand-drawn facsimiles of
the fragments, too. It is, therefore, vitally necessary that a close analysis of
the physical fragments themselves informs the researcher. Unfortunately,
the majority of scholarship on the Tel Dan fragments has proceeded on the
basis of the published photographs alone, and not on first-hand observation
of the fragments themselves.2
For this purpose, the epigraphical analysis of the three fragments was
conducted at the Israel Museum in Jerusalem where the fragments are
currently housed. Many hours were spent poring over the many physical
features of the actual fragments in great detail. As will be seen in the
presentation of this study, the results of this close and careful analysis
proved pivotal to a sound appreciation of the fragments and their meaning.
Indeed, many important aspects of the fragments have hitherto gone
undetected by many scholars for lack of such a careful analysis. I am con-
fident that this study will provide a coherent reappraisal and reinterpreta-
tion of the Tel Dan fragments.
2. See, e.g., V. Sasson, 'The Old Aramaic Inscription from Tell Dan: Philological,
Literary, and Historical Aspects', JSS4Q.I (1995), pp. 11-30 (13 n. 8).
Chapter 2
Introductory Remarks
For any inscriptional fragment it is important to first establish its archaeo-
logical context. That is, we must determine where and in what state or
usage the fragment was found. This information will prove invaluable for
determining the date of the fragment and then assessing its value in terms
of reconstructing a whole inscription and reconstructing the historical
background that gave rise to the inscription.
For the purposes of dating our three fragments from Tel Dan, we must
rely on the stratigraphy of the structural ruins at Tel Dan, as well as the
pottery assemblages found in the strata. The pottery remains will allow us
to date the strata and the various construction phases within these individ-
ual strata. The composition of structural walls will also aid us in discern-
ing phases. This will provide us with a chronological bracket in which to
place the fragments.
All three fragments were discovered in Area A of Tel Dan, located on
the southern slope of the mound. When the fragments were published, the
exact nature of the archaeological context for each was only partially
known. It is unfortunate that the conclusions of many regarding the dating
of the fragments were based on this partial knowledge. Indeed, Chapman
builds on the date for Fragment A proposed by Biran and Naveh and deter-
mines dates for Tel Dan's stratigraphy, rather than working the other way
around in a far safer and more methodical way. Chapman's reversed ap-
proach even leads him to reformulate the chronology of Iron II pottery.1
The dangers of pursuing such radical conclusions, or even more mundane
theories, before the archaeological context of an inscription is thoroughly
investigated is to 'jump the gun', so to speak. This can easily lead to dis-
1. R.L. Chapman, 'The Dan Stele and the Chronology of Levantine Iron Age Strati-
graphy', Bulletin of the Anglo-Israel Archaeological Society 13 (1993-94), pp. 23-29.
6 The Tel Dan Inscription
Fragment A
Fragment A (Fig. 2.1) was discovered on 21 July 1993, during the course
of regular seasonal excavations at the site of Tel Dan. The fragment was
located in the base of a wall (W.5073) that formed the western perimeter
of a complex of rooms (the location is marked 'A' in Fig. 2.2). Thus, the
fragment was found in secondary usage. This wall also marked the eastern
perimeter of a large paved piazza, approximately 400 m2 in area. These
constructions were all part of a massive gate complex at the southern
ramparts of Tel Dan, dated to Iron Age II.
Along the southern perimeter of the piazza, at the eastern extremity,
stood a gate. This is the outermost gate so far uncovered in this elaborate
complex. The find site of Fragment A was a little over 5 m north of this
2. N.P. Lemche and T.L. Thompson, 'Did Biran Kill David? The Bible in the
Light of Archaeology', JSOT64 (1994), pp. 3-22 (5).
3. B. Halpern, 'The Stela from Dan: Epigraphic and Historical Considerations',
BASOR 296 (1994), pp. 63-80.
2. Archaeological Context of the Fragments 1
Figure 2.1. Fragment A (Tel Dan Excavations, Hebrew Union College, Jerusalem;
photograph: Z. Radovan).
Figure 2.2. Plan of Structure A at the Tel Dan gate complex, indicating the locations at
which the Tel Dan fragments were uncovered (TelDan Excavations, Hebrew Union
College, Jerusalem).
8 The Tel Dan Inscription
threshold at the base of the perimeter wall.4 At the northwest corner of the
piazza, in the fortifications at the western perimeter, stood another gate.
Inside this gate was an enclosed courtyard with a canopied platform, as
well as a shrine to five massebot (sacred stones; sg. massebah). Legal pro-
ceedings undoubtedly took place in this enclosed area with the presiding
administrator seated on the canopied platform. Along the western wall of
this courtyard stood the main gate of Dan. A processional pavement led
from this gate to the west before turning north and reaching the city at the
top of the tel.
Quite carelessly, Biran and Naveh contradicted their information about
the exact location of the fragment when it was first published at the end of
1993. After carefully locating the fragment at the base of a wall (W.5073),
the publishers surmised at the end of their article that Ahab was likely to
have been the one who destroyed the original inscription and whose 'build-
ers reused a piece of it in the paving of the piazza'.5 This confusion over
whether the fragment was found at the base of a wall or as part of the pav-
ing of the piazza attracted much criticism from leading scholars.6 Any
doubt was dispelled with the publication of Fragments B1 and B2 in which
Biran and Naveh unequivocally stated that Fragment A was found 'in a
wall built on the flagstone pavement of the square'.7
The location of Fragment A, however, was not the only point that
attracted controversy. Biran and Naveh's dating of the fragment also un-
leashed criticism. According to Biran and Naveh, the layer of debris
covering the wall in which Fragment A was found was datable to Tiglath-
Pileser Ill's conquest of northern Israel in 733 BCE.8 This provided a sound
terminus ad quern for Fragment A. This point went undisputed and was
borne out by both the stratigraphy and pottery. The terminus a quo given
by Biran and Naveh was based on the pottery collected from the level
beneath the wall and piazza, which was dated to the first half of the ninth
century BCE.9 Biran and Naveh suggested that the original inscription had
been broken in the mid-ninth century BCE in accordance with the latest
pottery found in the lower level, thereby placing the terminus a quo in the
first half of the ninth century. This, however, does not equate well with the
stratigraphy of the area, a point noticed by Cryer10 and confirmed by later
reports of the complex east of the piazza.11
Cryer's objection was based on the fact that the wall in which Fragment
A was recycled must be younger than the paved piazza beneath it. Thus,
the original Tel Dan Inscription, while in its primary display position, was
contemporary with the piazza rather than the pottery found beneath it. The
original inscription, then, cannot have been destroyed at the same time as
the piazza was built. Rather, it was produced after the piazza was built
because it was not used as a flagstone in the paving, but as a part of the
younger wall.
These observations by Cryer do indeed call into question the dating of
the fragment's destruction proposed by Biran and Naveh. On the ceramic
evidence of the piazza alone, the mid-ninth century BCE is certainly too
early a date for the destruction and recycling of the Tel Dan fragments.
However, the stratigraphy of the building to the piazza's east is slightly
more complex than the equation drawn by Cryer.
A further factor to take into consideration here is the fact that much Iron
Age II pottery is not distinctive and, therefore, is only of limited value for
determining chronologies.12 Although this point was noted by Lemche and
Thompson, they do not hesitate to assign a date to the relevant stratum of
Tel Dan in the late eighth century BCE.13 Although the relative ambiguity
of some Iron II pottery is well known, in the case of Tel Dan there is
significantly less ambiguity. The reason for this stems from the well-docu-
mented invasion of the Galilee by Tiglath-Pileser III in 733 BCE, which
seems to have informed Lemche and Thompson's conclusion. Three
sources attest to this invasionnamely, the destruction layers in Upper
Galilean sites (such as Hazor and Tel Dan) and Transjordanian sites (such
14. Most notable among the texts are Layard 29b and III R 10,2 from Nimrud. For
further discussion, see J.K. J. Kuan, Neo-Assyrian Historical Inscriptions and Syria-
Palestine (Hong Kong: Alliance Bible Seminary, 1995), pp. 167-82.
15. That this destruction layer is not attributable to the earthquake of the mid-eighth
century BCE (see Amos 1.1) is demonstrated by the fact that only the gate complex at
Dan appears to have been damaged. This is consistent with military attack rather than
natural disaster.
16. I. Finkelstein, 'The Archaeology of the United Monarchy: An Alternative
View', Levant 28 (1996), pp. 177-87.
17. Biran has published some of the pottery found from these phases in 'Two
Bronze Plaques', Figs. 4, 7 and 11.
18. K.L. Noll, 'The God Who is Among the Danites', JSOT80 (1998), pp. 3-23 (4-
6, 19-20).
2. Archaeological Context of the Fragments 11
19. Halpern, 'The Stela from Dan', pp. 68-69. In his article, Halpern has mis-
takenly labelled the outermost gate as the 'inner gate'. This confusion undoubtedly
stems from the fact that the 'main gate' was further inside the complex, closer to the
city. Halpern seems to have understood the 'main gate' as being the outermost gate.
This, however, is not the case.
20. Halpern, 'The Stela from Dan', p. 69.
21. Biran, 'Two Bronze Plaques'.
22. Biran, 'Two Bronze Plaques', pp. 43-45, esp. Fig. 3.
12 The Tel Dan Inscription
century BCE, while the pottery beneath the floor was dated to the end of the
tenth century BCE.23 If we take Finkelstein's redating of Iron I strata, we
may bring these dates down slightly, so that the pottery on the floor of the
structure can be dated to the mid-ninth century BCE. This structure, termed
'Structure C', is therefore datable to approximately the second quarter of
the ninth century BCE, contemporary with the Omride Dynasty.24 On the
ceramic evidence, then, this structure is clearly earlier than the paved
piazza to the west. Biran suggested that the simple building in this phase
may have been used as a defensive outpost,25 a likely scenario since this
structure predates most of the gate complex in Area A.
The second phase of building saw the complex renovated and enlarged.
The western wall of the complex (W.5073) was extended 7 m northwards
during this phase. This extension was the thinner wall noticed by Halpern.
The complex itself consisted of three distinct chambers, at least one of
which had a paved floor. The pottery gleaned from this level dated to the
second half of the ninth century BCE,26 making it contemporary with the
paved piazza to the west. A scaraboid seal dated to the ninth century was
also found in one of the rooms,27 augmenting the date for this second
phase, termed 'Structure B'. Finkelstein's low chronology has virtually no
bearing on this date. We may assume, therefore, that Structure B, and the
paved piazza west of it belong to the same construction phase, begun in
the mid-ninth century BCE at the earliest. Thus, this phase may date to the
end of the Omride era or the beginning of Hazael's reign.
The portion of the western wall in which Fragment A was located was
not built until the third and final phase of building. In this phase, the
western wall was extended southwards and it was in this southward
extension of the wall that Fragment A was used as building material. This
phase is termed 'Structure A'. Cryer's observation that the stone's solid
composition and flat configuration led to its usage as a building material is
quite pertinent,28 particularly as the fragment was used as one of the base
stones of the wall. His statement that the wall in which Fragment A was
found is younger than the piazza adjacent to it, is correct only for this
extended portion of the wall built in the third phase of construction. The
pottery from this level dates to the first half of the eighth century BCE,29 a
date unaffected by Finkelstein's low chronology. This means that the origi-
nal inscription of which Fragment A was a part was broken just prior to
this phase of construction, some time at the end of the ninth century BCE
or, more probably, the beginning of the eighth century BCE.
We may surmise, therefore, that the original inscription was produced
sometime after the construction of Structure B and the paved piazza,
which occurred early in the second half of the ninth century BCE, but
before the third construction phase, represented by Structure A, in the
early eighth century BCE. The good condition in which Fragment A was
found suggests that the original inscription was destroyed before much
time had passed and before the stone had weathered any great deal. Thus,
from the archaeological context and the state of preservation of Fragment
A, we may surmise that the original inscription was produced towards the
end of this chronological bracket, and broken down not long after its
manufacture.
Fragment Bl
Fragment Bl (Fig. 2.3) was found on 20 June 1994, during excavations
along the base of Dan's Iron Age wall. After a Roman clay pipe from a
later stratum had been removed from this area, a small paved platform was
found, 4.5 x 2.5 m in dimension. This platform had served as a shrine to
three massebot found on the northern edge of the platform. The three mas-
sebot were arranged along this edge from left to right, largest to smallest.
Directly in front of the largest massebot which stood 117 cm tall, was a flat
base on which a decorated capital stood. A basalt bowl with traces of fire
inside it was found on this capital. Traces of ash around the stones con-
firmed the platform's use as a cultic shrine at which offerings were made.
This small shrine was built over debris dated to Tiglath-Pileser Ill's
conquest of northern Israel in 733 BCE. This debris also covered Structure
A, which represented the third phase of construction of the complex on the
eastern side of the paved piazza. In the course of clearing this debris, Frag-
ment B1 was discovered approximately 2 m south of the shrine's platform
(location 'Bl' in Fig. 2.2). The fact that Fragment Bl was found in this
debris suggests that it came from a later context than that of Fragment A,
29. Biran, 'Two Bronze Plaques', pp. 50-51, esp. Fig. 11.
14 The Tel Dan Inscription
which had been covered by this debris some 13m to the southwest.
However, this view had to be altered with the discovery of Fragment B2
just ten days later.
Figure 2.3. Fragment Bl (Tel Dan Excavations, Hebrew Union College, Jerusalem;
photograph: Z. Radovan).
Fragment B2
As excavations along the base of the Iron Age city wall continued east-
ward, five massebot were uncovered at the base of the wall. A probe
beneath the two easternmost massebot revealed that they had been laid
prior to the paving of the area with flagstones. The base of the massebot
was approximately 40 cm below the level of this pavement. The pottery
gleaned from beneath the flagstones dated to the end of the ninth century
BCE and the beginning of the eighth century BCE. Thus, this pavement at
the base of the wall, to the north and east of Structure A, was constructed
in the early eighth century BCE, probably during the Israelite revival under
Jehoash and his son, Jeroboam II. We may reasonably surmise, therefore,
that this area was paved at the same time that Structure A was built. There-
fore, this area to the north and east of Structure A is to be dated slightly
later than the paved piazza west of Structure A. The piazza was contem-
porary with Structure B (early second half of the ninth century BCE), while
the pavement along the city wall was contemporary with Structure A
(early eighth century BCE).
On 30 June 1994, Fragment B2 (Fig. 2.4) was discovered as a constitu-
ent part of this pavement, contemporary with Structure A. The builders of
2. Archaeological Context of the Fragments 15
On the basis of this synchronism, we can date the fracturing of all three
fragments to the early eighth century BCE. This suggests that there is some
relation between all three fragments, even if it is only chronological in
nature. In other words, on the basis of the archaeological evidence, we can
say with some degree of certainty that all three fragments were broken at
the same time. The subsequent recycling of Fragment A and Fragment B2
as building matter should be dated to the revival experienced in Israel
during the reigns of Jehoash (c. 798-782 BCE) and Jeroboam II (c. 782-
753 BCE).30 The pottery assemblage found in the levels of this construction
phase spanning the first half of the eighth century BCE means we should
assign this construction phase to the reign of Jehoash (c. 798-782 BCE).
We can also surmise that Fragment B1 was also recycled in some way
during this construction phase, though the structure in which it was used
has not been preserved. This synchronic parallel between all the fragments
strengthens the connection between them. Only Fragments B1 and B2 (col-
lectively 'Fragment B') are clearly from the same inscription by virtue of
the join along a common fracture line. The archaeological context does
not, however, inform us as to whether Fragment A was also part of that
same inscription. What is clear, though, is that all three fragments are from
the same era.
Chapter 3
EPIGRAPHICAL ANALYSIS
Introductory Remarks
In the case of the Tel Dan fragments, epigraphical analysis is of paramount
importance. The issue of whether we have parts of one or two inscriptions
must be clarified if we are to have any hope of a sound historical appraisal.
A comprehensive epigraphic analysis of what we actually have will do
much to further our chances of assigning at least a modicum of meaning to
the fragments.
In this part of my analysis, I will make close and careful observations
regarding the physical fragments themselves, and the nature of the individ-
ual incised strokes, and so decipher each individual grapheme. Since there
is no doubt that Fragments B1 and B2 make up an integral unit, they will
be dealt with collectively as 'Fragment B'.
stone. This original edge, however, has undergone some slight erosion so
that it has dulled and become somewhat rounded. There are also clear
signs of scarification along the edge. The random nature of these lacera-
tions suggests that they were sustained during either breakage of the origi-
nal inscription, transportation, or simply as a result of wear and tear while
on display or in secondary usage.
A portion of the original side of the inscription has survived along this
particular edge of Fragment A. Here, the stone has been obviously wrought
smooth, though the surface is punctuated by abrasions and numerous pock
marks. The surface on which the writing is displayed is also smooth. A
very slight convex curvature of the written face is detectable when the frag-
ment is viewed from the side. It is difficult to say whether this curvature
along the face of the inscription was intentional since it is of such subtlety.
The smoothness of both the written surface and the right-hand side prompts
the logical conclusion that the original inscription was professionally fash-
ioned and smoothed. Of itself, this implies the inscription was of quite
some significance.
The fashioned smoothness of the fragment's right-hand side is also
critical for reconstructing the original appearance of the inscription and
how it was displayed. There are two possible reasons for why the side was
smoothed:
1. Structuralthe inscription was intended to be displayed in a
wall. Thus, the stone on which the inscription appeared had to be
uniformly shaped to sit evenly as an ashlar block within the struc-
ture of a wall; or
2. Aestheticthe inscription was intended to be a free-standing
stele, much like that of the Mesha Stele from Dhlbn in Moab.
As such, the stone had to be groomed in order to be aesthetically
pleasing from all angles.3 To this end, not just the written face
and right-hand side would have been smoothed, but all sides,
including the bottom.4
In order to assess which of these two possibilities is the more probable,
we must consider the stone's state of wear, as well as the wider archae-
3. Unlike the Mesha Stele, however, Fragment A does not have a rim bordering
the written surface.
4. The bottom face would have rested on the ground or on a plinth of some sort.
Despite not being visible, it would presumably have been smoothed for both structural
and aesthetic reasons. Compare the structure of the Mesha Stele.
20 The Tel Dan Inscription
Figure 3.1. Facsimile of Fragment A highlighting area of erosion on the right side
of the written surface(drawn by Ada Yardeni).
The oddity of this belt of erosion is that it is the only sign of consistent
natural weathering on the written surface of Fragment A. Similar areas of
erosion appear in other places across the written surface of Fragment A,
but they are so much smaller in size that we can only call them patches.
Also, the depth of erosion in most of these patches is far less than that of
3. Epigraphical Analysis 21
the large eroded belt on the right-hand side of the fragment. The belt of
erosion is also more regular than the other smaller patches of erosion.
This means that this narrow area along the edge of the fragment was the
only part of the written surface exposed to the elements. Schniedewind
observes that this edge appears to narrow at the top of the fragment, sug-
gesting that the original whole inscription was rounded at the top in a
similar way to the Mesha Stele.5 However, a closer examination of the
stone at this point merely shows that there has been slightly more erosion
of the edge in this upper portion than in the lower portions. There is, in
fact, no sign that the edge was deliberately rounded. This does not, of
course, preclude the possibility that the original inscription was rounded at
the top. It is just that the portion of the inscription represented by Frag-
ment A shows no signs of this.
The incisions making the letters were quite stylistically done. A very
close analysis of them reveals that the actual channels of incision are, for
the vast majority, quite smooth and usually rounded (see Fig. 3.2). For
Biran, this prompted the response that the engraving instrument, which we
will label a 'chisel', had a round edge,6 presumably like the end of a match-
stick. However, such an instrument would be most curious since chisels
and other such engraving tools are made with sharp edges to promote
cutting. A round-edged chisel or 'stylus', as Biran terms it, is not an effec-
tive cutting tool, and the expert finish of the inscription suggests that fine
quality tools were used. As such, the nib of the chisel was likely to have
had a sharp edge, much like a flat-bladed screwdriver. Indeed, in experi-
ments conducted by the author, such a tool inevitably produced incisions
on stone that were comparable to those on Fragment A.
Biran also deduces that the engraving tool was probably made of iron.7
Certainly the material could not have been flint or stone since such mate-
rials are not given to being shaped into long, thin tools which are strong
enough to withstand hammering and carving into stone.8 Tel Dan definitely
5. W.M. Schniedewind, 'Tel Dan Stela: New Light on Aramaic and Jehu's Revolt',
BASOR 302 (1996), pp. 75-90 (78); W.M. Schniedewind and B. Zuckerman, 'A Possi-
ble Reconstruction of the Name of Haza'el's Father in the Tel Dan Inscription', IEJ
51.1(2001), pp. 88-91(88-89).
6. Biran and Naveh, 'An Aramaic Stele Fragment', p. 85.
7. Biran and Naveh, 'An Aramaic Stele Fragment', p. 85.
8. Knives made of flint, although fitting the description of being a long, thin tool,
were not used to carve or chisel stylistic incisions into stone.
22 The Tel Dan Inscription
9. A. Biran, Biblical Dan (Jerusalem: Israel Exploration Society, 1994), pp. 147-
57. Biran dates these fixtures to the early tenth century BCE. I have adjusted this date
downwards in light of the arguments proposed for a low chronology by Finkelstein in
The Archaeology of the United Monarchy'.
10. Thanks go to Sariel Shalev of the Weizmann Institute of Science, who con-
ducted mineral analyses of the metal slags found at these installations, and who
provided this information.
11. Thanks go to David Ilan of Tel Aviv University for this information.
3. Epigraphical Analysis 23
against the fragments by Cryer.12 Since the argument for this claim rests
on evidence from the surfaces exposed in breakage, we will need to exam-
ine these surfaces as closely as the written surface.
12. The first charges of forgery were made against Fragment A by G. Garbini in
'L'iscrizione aramaica di Tel Dan\Atti della Accademia nazionale deiLincei, Scienze
morali, storiche efolologiche, rendiconti 9.5.3 (1994), pp. 461-71. Cryer then raised
questions as to the authenticity of Fragment A in his article, 'On the Recently Discov-
ered "House of David" Inscription', though he stated that he did not believe the frag-
ment was a forgery. However, at some stage after the publication of Fragment B, Cryer
changed his opinion and regarded all three fragments as forgeries. Cryer privately com-
municated to me the nature of the evidence that led him to this conclusion. This evi-
dence will be dealt with below. I am indebted to Cryer for supplying this information.
13. The author conducted trials in which strokes of various shapes were chiselled
onto stone surfaces of varying types from rough, unhewn sandstone to smooth slabs of
concrete. The chisel was also held with many different grips. Invariably, the starting
point of each incision made was notably shallow, while the end point was character-
ized by a deep indentation.
14. This is true even after factoring in the patch of erosion in this area of the in-
scribed surface.
24 The Tel Dan Inscription
Figure 3.3. Deep and shallow points on the incisions indicating the starting point and
direction of engravingfor the individual letter strokes (Tel Dan Excavations, Hebrew
Union College, Jerusalem; photograph: Z. Radovan [detail]).
Figure 3.4. Illustration showing the direction in which individual letter strokes were
inscribed (detail from facsimile drawn by Ada Yardeni).
15. This assumption has been made purely on the basis of probability. In our day,
the great majority of people are right-handed. In the text of Judg. 3, Ehud is noted as a
left-handed man in a way that suggests that left-handedness was uncommon. Therefore,
we can assume that in antiquity, just as many, if not more, people were right-handed.
26 The Tel Dan Inscription
move towards the right. This would, indeed, be the most natural order of
carving the letters if the engraver was situated along the left edge of the
inscription.
This particular position of the engraver in relation to the stone also tells
us that the inscription was certainly written and engraved prior to being
placed in its final display position, not afterwards. Indeed, it overwhelm-
ingly points to a scenario in which the stone of the inscription was laid flat
with the written surface facing upwards as the craftsman engraved the
letters. It is fair to presume that the inscription was eventually displayed
with the written surface perpendicular to the ground.16 If the inscription
had actually been in this display position when the letters were being
carved, we would expect different data in regards to the direction of carv-
ing and the order of strokes carved.17
A further implication of knowing the position of the engraver is that the
text of the inscription must have been chalked or marked out before any
chisel was put to the stone. It is an implausible suggestion that the engraver
carved the inscription straight onto the stone without some sort of guide or
stencil. Indeed, it is difficult to imagine the engraver composing the inscrip-
tion while carving it from the stone's left edgea position that would not
afford him readability. Rather, the fact that the engraver seems to have
worked from a flanking edge of the inscriptions means that he need not
have been literate at all. Since literacy was very limited in antiquity,18 it is
plausible that a scribe, specially commissioned with the composition of the
inscription, chalked the text onto the stone. The engraver would then have
traced out with a chisel what to him were not necessarily legible letters,
but simply strokes and curves marked on a stone.
16. No other monumental lapidary inscriptions seem to have been displayed in any
other way.
17. E.g., we would probably expect the vertical strokes to have been generally
carved in a downward direction as opposed to upwards. We would also expect
horizontal strokes to have been carved from right to left, as opposed to left to right.
18. Harris estimates that only 10 per cent of the Athenian population in the fifth-
fourth centuries BCE were literate and that this was an exceptionally high rate because
of the nature of Athenian society. See W.V. Harris, A ncient Literacy (Cambridge, MA:
Harvard University Press, 1989), p. 114. We expect that the rate in other earlier soci-
eties was much less. See also D.W. Jamieson-Drake, Scribes and Schools in Monar-
chic Judah: A Socio-Archaeological Approach (JSOTSup, 155; The Social World
of Biblical Antiquity, 9; Sheffield: Almond Press, 1991); I.M. Young, 'Israelite Liter-
acy: Interpreting the Evidence. Part I', PT48 (1998), pp. 239-53; idem, 'Israelite Lit-
eracy: Interpreting the Evidence. Part IF, FT 48 (1998), pp. 408-22.
3. Epigraphical Analysis 27
Fragment A Degrees:
+ above the horizontal
below the horizontal
Line 1 -11+
Line 2 -10+
Line 3 -8
Line 4 -5
Line 5 -4
Line 6 -3
Line 7 -2
Line 8 0
Line 9 0
Line 10 0
Line 1 1 0
Line 12 +2
Line 13 +4
19. These figures are the average angles of the lines between their extant starting
and finishing points. It should be noted that there are variations of angles within each
line. For example, Line 3 starts seemingly straight but then slants down after the first
word, IO2H. Also, in Line 5, the slope of the word "Tin is straighter than every other
word in that line.
28 The Tel Dan Inscription
These figures demonstrate the very real trend of the lines coming closer
to the horizontal further down the inscription, with the possibility that they
begin to slope upwards in the last two lines of the fragment.20
When this information is combined with that of the angles of each
incision of each letter, a picture emerges of the letters being gradually
rotated clockwise throughout the inscription. That is, after factoring in the
slope of each line of text, the letters which appear in the top portions of
Fragment A have a slight leftward slant, but in the bottom portions, they
have straightened up.
Since the data point to the engraver being positioned at the left edge of
the inscription during carving, it is unlikely that the slanted lines and let-
ters in the top part of Fragment A are attributable to the process of engrav-
ing. First, we have established that it is highly unlikely the engraver wrote
the original text as he engraved it since his stance did not afford readabil-
ity and comprehension of the text. Second, even if the engraver could have
produced the original inscription from this position, no posture could have
made him slant the lines as they have been. As such, if the slanted lines
and letters are attributable to the craftsman during engraving, then it was a
purely random force that exerted itself on him which caused this result.
However, the figures of the line slants indicate that it was not a random
factor at work herethere is a certifiable trend for the lines to become
more horizontal, and as such, there must have been a consistent factor
behind it.
As has been suggested, it is likely that the text of the inscription was
originally chalked onto the writing surface and subsequently etched out by
an engraver. With this suggestion, we can posit a consistent factor as a
cause of the slanted lines and letters. Writing is logically carried out in one
posture, namely, one that not only allows the scribe to write text, but also
to be able to read what he is writing. We have confirmed that the stone of
the inscription was originally laid flat with the reverse side on the ground.
Therefore, the scribe must have been positioned at the bottom edge of the
inscription when he wrote it, and would have been limited to this spot in
order to be able to write and read the text. Since the extant Line 1 is almost
certainly not the first line of the original inscription (see discussion of Line
1 below), the scribe must have written at least one or two lines of text
above the extant Line 1. Also, the extant Line 13 is almost certainly not
20. Since the slopes of Line 12 and Line 13 are statistically ambiguous, this remains
only a possibility that cannot be confirmed.
3. Epigraphical Analysis 29
the final line of the original inscription.21 Thus, the original inscription
would have stood taller, and with more lines of text, than the current sam-
ple from Fragment A.
With these considerations, we can then note that the lines slope away to
the left in a way that is consistent with writing from right to left with an
outstretched arm and without the aid of ruled lines. In other words, it
appears that the scribe who first chalked the inscription onto the stone had
to stretch his arm forward in order to write the upper portions of the origi-
nal text, thus limiting his freedom and ease of movement. With this posture
and the lack of marked lines to keep the text level, the text naturally sloped
downwards to the left.
This is also evidenced by the somewhat cramped nature of the letters in
these upper portions of Fragment A, and the gradual improvement of gen-
eral neatness towards the bottom of the fragment. This is consistent with a
gradual freeing of movement and improved precision in writing as the arm
is stretched less and brought closer towards the body further down the
inscription. In short, the steadiness of the hand increases and less volatility
is likely in the script.
We should ask whether any other factors could have caused the improve-
ment in the inscription's neatness. However, when we do, we find no
theory offering a consistent explanation for the trend visible on Fragment
A. It is unlikely that the disparities in both line slopes and letter neatness
are the result of multiple hands; that is, that more than one scribe chalked
the text onto the stone. If this had been the case, we would expect a more
abrupt change in the appearance of the letters since there would be one, or
possibly more, definite transition points. However, as the appearance of
the letters changes gradually throughout the inscription, oscillating at times
between steep and flat but generally becoming flatter towards the end, the
theory of posture is a more likely explanation.
It could perhaps be argued that the difference in multiple hands was
essentially eradicated by the conventions of one engraver. However, the
consistent improvement in line slopes still indicates that one hand is behind
the writing of the text, and that one consistent, enduring and natural force
resulted in the steady improvement of scribal quality. Similarly, this last
21. The clean breakage at the bottom of Fragment A suggests that the piece of
stone which was broken here (and which is no longer extant), was heavy in weight. As
such, we can expect this piece to have had numerous lines of text. The content of Line
13 also suggests that there was considerably more text after it. This will be discussed at
a later point.
30 The Tel Dan Inscription
factor undermines any suggestion that the scribe was initially careless with
the straightness of his lines but was more careful in the lower lines of the
text. In fact, all things being equal, a scribe is more likely to become more
careless rather than more cautious throughout the writing of a text.
Having gleaned such information about the writing and carving of the
inscription, we must also mention the limitations of our evidence, though
these are of minor consequence. Though we can calculate where the en-
graver stood, the direction and order of the strokes in which he carved, and
in what position the stone was during carving, we have no way of telling
whether the engraver started carving the top lines of the inscription first, or
the bottom lines. As has been noted, to an engraver, the order in which the
figures and lines of text are carved is not dependent upon how the text is
read; it is not the written text that matters, but the incising of individual
strokes. Thus, there would have been no restriction as to which lines of
text were to be carved first and which last. Since close examination of the
carved strokes has revealed that the engraver worked individual letters
from left to right, it is only reasonable to assume that entire rows of text
were worked from left to right also (or, from the engraver's point of view,
from bottom to top). However, we have no way of knowing whether the
lines of text were carved in their rows, or whether individual letters on
successive lines were carved. Neither do we know whether the engraver
worked from Line 1 down to Line 13, or in the opposite direction, or in a
completely random order.
22. This is the average length of a modern man's arm. Ancient people were slightly
smaller in size than the average person today, so the calculation is probably closer to
60 cm than 70 cm. Skeletal remains found at various sites in the Mediterranean basin
3. Epigraphical Analysis 31
above the original bottom edge of the whole inscription.23 We can average
this out to 65 cm.
Furthermore, there is a length of 15 cm of the written surface above
Line 8 that can be added to this approximate length. However, the newly
exposed parts of the stone indicate that Line 1 of Fragment A was not the
uppermost line of the original inscription. Thus, we may confidently add
the extra 5 cm of the fragment's height above Line 1 to our estimations.
These extra 20 cm yield us a minimum approximate height of 85 cm.
Yet, we must also factor in the possibility, indeed the likelihood, that the
scribe did not only stretch his hand out over the stone, but that he also
leaned part of his body over the stone as he wrote. It would be peculiar if
the scribe could have rectified the line slopes and neatness of his letters by
just leaning forward a little and yet failed to do so. As such, he almost
certainly leaned part of his body forward over the stone as he wrote (see
Fig. 3.5). This would add a considerable amount to the length of stone. I
propose adding a further 25 cm to the height of Line 8 in addition to the 65
cm already calculated above. These extra 25 cm are in accordance with the
results of simple experimentation with this posture.24 Thus, I posit Line 8
being approximately 90 cm above the bottom edge of the inscription. This
gives us 110 cm as the approximate minimum height of the original inscrip-
tion stone when it stood in tact. Since neither Fragment A nor Fragment
B1 show any trace of the inscription's original top edge, there is no way of
suggesting a maximum height for the inscription.
attest that the average adult male was approximately 165 cm tall, factoring in shrink-
age. See, e.g., U. Gaus, V. Von Grave and M. Kerschner, 'Milet 1990: Vorbericht iiber
die Arbeiten des Jahres 1990', IstMitt+++++++++++++++++++++
23. This, of course, reasonably assumes that the bottom edge was very close to, or
touching, the scribe's body.
24. The author conducted a very simple experiment in which the arm was extended
fully forward (with a pencil in hand) and the body leaned forward over a writing
surface, with the writing surface being at the level of the waist. The distance was then
measured between the edge of the writing surface at the waist and the point on the
writing surface over which the shoulder of the extended arm was situated. This yielded
a result of 28 cm. This was then rounded down to the nearest significant integer on the
basis that human remains from the Iron Age indicate that adult males were of smaller
and shorter build than adult males today.
32 The Tel Dan Inscription
Figure 3.5. Illustrations demonstrating the scribe's outstretched arm and forward-
leaning posture during plotting of the upper portions of the text (left) and the more
comfortable posture attained during plotting of the lower portions of text onto the
inscription surface (right).
It must be noted, however, that this estimation is for the height of the
stone itselfnot the area of actual written text. We must consider the
possibility that text did not cover the full length of the stone. The Mesha
Stele, which measures a comparable 115 cm in height, originally had a
base or plinth carved into the stone as a relief.25 A similar convention may
have been employed for the Tel Dan Inscription. Similarly, we must also
mention the possibility of pictorial reliefs carved into the stone either
above, below, or to the side of the text. Though nothing of the sort has
survived among the fragments, we have neither the uppermost, lowermost,
or leftmost portions of the original inscription. This fact, coupled with the
relatively small sample of the whole inscription that has survived means
we must not discount this possibility. The plausibility of such pictorial
reliefs is demonstrated by other inscriptions, such as the stele of Kilamuwa
and the Nerab Stelae.
As we turn to the task of deriving an approximate width for the inscrip-
tion we find that none of the three fragments discovered thus far preserves
the original left edge of the inscription. However, we can make certain
observations about the individual engraved strokes and the posture of the
engraver to arrive at an estimate.
25. This relief base was destroyed when the Mesha Stele was sabotaged in the
nineteenth century. For an account of the drama, see A. Lemaire, ' "House of David"
Restored in Moabite Inscription', BARev20.3 (1994), pp. 30-37 (31-34). The base has,
however, been reconstructed for display purposes. The inscription is currently kept at
the Louvre Museum in Paris.
3. Epigraphical Analysis 33
First, as was noted previously, the individual letter strokes point towards
the engraver having been situated along the++++++++++++++++++++++
will see that this position is further confirmed by the evidence of all the
other letters on Fragment A, including those on the rightmost edge of the
fragment. Thus, taking the most likely view that the engraver was right-
handed, the strokes demonstrate that the engraver's position allowed him
to carve the letters on the far side of the inscription (along the right edge)
with relative ease and precision. As such, the inscription's width at the
latitudes represented by Fragment A was no more than 105 cm across, in
keeping with the results of a posture extended fully forward, as was seen
in discussion of the scribe's posture.
However, it was noted that the scribe demonstrated some difficulty in
writing the text of the inscription due to this posture. Whereas a scribe is
limited to one position in relation to the stone, namely, a position that
affords readability, an engraver is free to move about the stone. He is not
limited to a position that affords readability of the inscriptionjust a
position that affords visibility of individual strokes composing the letters,
and comfort in engraving these strokes with a chisel and mallet. Such posi-
tions exist all around the perimeter of the writing surface. Thus, it is telling
that the engraver inscribed all the letters of Fragment A from only one
edge of the perimeter, namely the left edge (see Fig. 3.6). It tells us that
from this one edge there was no discomfort or lack of precision during
carving, especially considering the fine quality of the carving (as opposed
to the handwriting of the scribe).
of usage as building material for a wall. The shape of the recess is far too
regular for this to have been the case. It is clearly the result of deliberate
human activity.
Second, crucial to our analysis of this recess is the micro-fragment
inside it, which contains portions of letters belonging to Line 8 and Line 9.
The micro-fragment fits neatly along the line of breakage that runs down
the left side of Fragment A from Line 5 to the bottom of the fragment.
This tells us that the micro-fragment was fractured at the same time this
large break was sustained. Furthermore, since the micro-fragment is also
an inherent part of the recess, of which only half is on Fragment A, the
recess and line of breakage must also have been sustained simultaneously.
The best explanation accounting for these facts is that some kind of instru-
ment was used on the inscription to destroy it, perhaps a hammer, mallet
or other such piece of equipment. The blow of this hammer caused this
area of the stone to fracture and in the subsequent fall of the inscription,
some of the fractures were widened to complete breakages.
Presumably, the hammer was struck against a central position of the
inscription. This is only to be expected for an act of deliberate maximum
damage. This being the case, the recess would represent the approximate
centre in the width of the total inscription. In keeping with the action of
swinging a hammer, the recess also probably represents the approximate
height of a man's elbow or waist: about 85-90 cm. This matches perfectly
with the calculated height of Line 8, which intersects the recess.
Thus, measuring from the estimated centre of this blow to the right edge
of the inscription, a figure of 17 cm is returned, and hence a total approxi-
mate width of 35 cm. Such a figure is within the suggested maximum
width of about 45 cm (a cubit). This correspondence between the figures
suggests that the proposed theory on the cause of the recess is indeed
correct. Therefore, the calculated dimensions for the written surface are
approximately 110x35 cm. We must remember, however, that 110 cm is
a minimum height. The true height of the original inscription is unlikely to
have been much more than that. References to the author's father are
likely to have been made at the beginning of the inscription rather than
later. As a result, the first few lines of Fragment A are probably within the
first few lines of the original inscription.
Line 1
The first line extant on Fragment A preserves only four letters. By com-
parison with better preserved lines (e.g. Line 5), it is calculated that there
36 The Tel Dan Inscription
is room for eight, or possibly nine, letters before the first extant letter on
Line 1. The breakage of the inscription, however, has lost these letters to
us. It is doubtful whether Line 1 is the original first line of the whole
inscription because the stone is large enough to have had at least one or
two lines of text before Line 1. This is evident after considering the height
of the newly exposed29 face at the upper left portion of the fragment since
it is the exact same space required for two lines of text (see Fig. 3.7).
Figure 3.7. Illustration showing enough roomfor at least two lines of text before Line 1
of Fragment A (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph:
Z. Radovan [detail]).
Biran and Naveh understood the first visible strokes at the extant begin-
ning of Line 1 as forming the figure of a mem.30 As such, the long vertical
stroke was understood as the stem of the mem, joining the rest of the figure
at a point above the line of breakage. At a cursory glance, this seems plau-
sible, even probable. However, we must find strong objection to it. Tropper
was first to note the possibility that the jagged construction understood by
Biran and Naveh as the head of the mem, may actually be a sin.31 As such,
Tropper understood the preceding vertical stroke as the remains of a
probable taw. This possibility was noted by Schniedewind, who rejected it
on the basis that the taw and sin would be too close together.32 However,
such is not the case. At the end of Line 8, we see the remains of a taw and
lamed occurring much closer together than the vertical stroke and jagged
construction here in Line 1. Indeed, there is far less space between the let-
ters in Line 8 than there is here in Line 1. Thus, there is ample evidence to
allow the reconstruction of a taw and sin here.
29. That is, newly exposed when the original whole inscription was broken.
30. Biran and Naveh, 'An Aramaic Stele Fragment', pp. 87, 90.
31. J. Tropper, 'Bine altaramaische Steleninschrift aus Dan', UF'25 (1993), pp.
395-406 (401-402).
32. Schniedewind, 'Tel Dan Stela', p. 79.
3. Epigraphical Analysi+ ++
Yet, despite the room for this reconstruction, we must test to see if it
actually is the case. Schniedewind's initial observations are invaluable
here, though I am surprised that he rejected their import. Schniedewind
noted the difference in slant between the strokes of a sin and the strokes on
the head of a mem. The key stroke to consider is the leftmost stroke of the
figure. On all the attestations of mem in Fragment A, the leftmost stroke
leans to the right. On the other hand, all the attestations of sin invariably
show the leftmost stroke leaning decidedly to the left. When we examine
the leftmost stroke of the figure here in Line 1, we see that it too leans
markedly to the left. Since the evidence is unanimous that such a trend
indicates a sin, we must inevitably conclude the same for this figure. This
is also confirmed by closely comparing the long vertical stroke which
Biran and Naveh understood as the stem of a mem with the next stroke
which Biran and Naveh thought joined it. These strokes are seen to be
parallel and show no signs whatsoever of converging above the line of
breakage. Thus, we must understand the jagged construction as a sin.
This leaves us with the vertical stroke preceding this sin. Tropper's
suggestion of taw is the most probable because of the space involved. The
only other letters that show a comparably angled stem are kaph, mem, nun
anpeh. A mem is automatically ruled out for lack of space. Similarly, all
instances ofkaph, nun andpeh in Fragment A (and Fragment B) are too
large to fit in this narrow space. We have no such problems with accom-
modating a taw. Therefore, Tropper's suggestion of a taw followed by sin
at the beginning of Line 1 is to be retained.
The incisions of both the taw and sin are clearly not as deep as those of
most other letters. This can be attributed to two factors. First, it appears
that the engraver did not chisel these letters very deeply,33 and second,
there is a slight patch of erosion that covers most of these letters. A slight
chip is also visible along the shallow stem of the taw. This chip is quite
anomalous to the formation of the letter since it has obliterated part of the
incised stem. As such, it cannot have been on the stone's prepared surface
before the inscription was engraved. Neither can it be the result of an
accidental slip of the engraver's chisel. Rather, it must have been sustained
after the engraving. However, this chip has also suffered some erosion.
Thus, the chip was made before the erosion in this area of the stone
occurred.
33. We can only speculate as to why this was the case. Of course, it is impossible
without modern machinery to keep the depth of incisions exactly the same. As such,
the shallowness of this letter might be partly attributable to chance.
38 The Tel Dan Inscription
The question we must then ask is when exactly the chip was made. Was
it inflicted on the stone when it was in its original display position, or
during the breakage of the stone and subsequent recycling as building
material? Two similar chips further to the left provide a clue that will help
us answer this question. Figure 3.8 shows the position of these two chips
in relation to the first chip on the stem of the letter taw.34 They are on the
approximate same latitude as this first chip and have also undergone
similar erosion. All three chips are of a similar nature, being wide and flat
in shape. Their proximity to each other and similar characteristics suggest
they have a similar or even common cause.
Figure 3.8. Three chips below Line 1 with the area of abrasion caused by breakage of
the stone (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z.
Radovan [detail]).
The leftmost chip ('Chip 3' in Fig. 3.8) provides the vital clue. At the
leftmost extremity of this chip, we observe an abrasion that is clearly asso-
ciated with the breakage of the stone; it continues over the broken edge
onto the newly exposed face, as well as further down to intrude upon a
partly destroyed letter in Line 2 (see Fig. 3.8). The leftmost point of Chip
34. Curiously, the facsimile of Fragment A drawn by Ada Yardeni and published
by Biran and Naveh on p. 89 of 'An Aramaic Stele Fragment', does not detail these
three chips. Neither were they shown when the facsimile of Fragment A was coupled
with a facsimile of the two joined B fragments in Biran and Naveh, 'The Tel Dan
Inscription', p. 12.
3. Epigraphical Analysis 39
3 appears to have been scored into this abrasion, and must therefore have
been sustained after the inscription was broken. Since Chip 1 and Chip 2
bear the same characteristics as Chip 3, we can reasonably deduce that
these too were made after the stone was broken. As a result, we can con-
clude that the patch of erosion which covers the first two letters, taw and
sin, and a small area around it was not sustained while the inscription was
in its original display position, but after the inscription was broken. Subse-
quently, the three chips were sustained during the transfer and reuse of the
fragment as building material.
After the taw and sin we see a fully preserved resh. In contrast to the
preceding two letters, the incisions of this resh are quite deep. In fact, so
notably deep are they that a small part of the stone within the area of the
triangular head has been chipped off during the chiselling of the lines. This
chip runs along the length of the left vertical incision such that it appears
to be part of the incision, though not quite in harmony with it. It is not
anomalous to the incised stroke, unlike the chip along the stem of the taw
two letters before. However, like the taw, the stem of the resh is quite
shallow in depth due to the small patch of erosion in this area of the stone.
Following the resh on Line 1 is a word divider. It is in the shape of a
simple dot and has been carved extremely well. Indeed, it presents to the
reader a very neat circle and the indentation into the stone itself is almost
perfectly hemispherical. This suggests that the engraver first chiselled a
rough mark, which he then ground out, probably without the use of a
hammer, giving the neat circular effect.
The final letter extant on Line 1 is an 'ayin. The letter is almost fully
preserved with only the very left side of the figure damaged in the in-
scription's breakage. No other incised strokes give any indication that this
is possibly a qoph, and the possibility of a lamed is dispelled by the fact
that the letter appears to be a closed circle.
We may decipher the inscribed characters on Line 1, therefore, as:
[...]unoi[...].
Line 2
The beginning of Line 2 is very difficult to decipher. The broken edge
where the written surface meets one of the newly exposed faces (the upper-
most) has been significantly abraded such that there is no clear distinction
between where the written face ends and the newly exposed face begins
(see Fig. 3.9). This poses significant problems for deciphering the first few
letters of Line 2 as the abraded edge cuts through them. Consequently, the
40 The Tel Dan Inscription
first legible letters of Line 2 are quite a way along the line. The first clearly
legible character in Line 2 is a word divider. Like the word divider in Line
1, it is a very round figure that appears to have been hollowed out after it
was initially chipped into the stone.
Figure 3.9. The abraded edge between the -writtenface and the uppermost exposed face
(Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan
[detail]).
However, there are still some traces of the damaged letter immediately
preceding this word divider. We must consider whether this letter, and
indeed any letter preceding it on Line 2, can be reconstructed. In neither of
their two publications do Biran and Naveh offer any suggestion for these
letters. Yardeni's facsimile of Fragment A offers no reconstruction either,
though she shows that some unclear markings are present on the face of
the stone.35 A careful examination of the actual fragment shows what
appears to be the bottom left portion of a letter etched into the stone. The
visible remnant of this damaged letter is fairly rounded. It does not appear
to have a low stem or tail as the written surface is clearly legible where we
would expect it to be. This being the case, we cannot consider beth, heh,
kaph, mem, nun, samekh,peh, sadhe, qoph, resh or taw as possibilities for
restoring the letter. Therefore, in order to reconstruct the letter, we must
look for letter shapes that display roundness in the bottom left portion and
which have no low stem. As such, we may suggest teth, yodh, lamed and
'ayin, each of which must be dealt with on their own merits.
The difficulty in proposing a teth here is that neither Fragment A nor
Fragment B contain any examples ofteths.Thus, we have to turn to the
35. Biran and Naveh, 'An Aramaic Stele Fragment', p. 89, Fig. 8.
3. Epigraphical Analysis 41
36. For the text of the Amman Citadel Inscription and an analysis, see S.H. Horn,
'The Amman Citadel Inscription', BASOR 193 (1969), pp. 2-19.
37. The definitive edition of the Sefire Treaties is J.A. Fitzmyer, The Aramaic
Inscriptions of Sefire+++++++++++++++++++++++++++++++++++++
38. For the text and facsimile of the Mesha Stele as well as a treatment of
numerous issues arising from it, see A. Dearman (ed.), Studies in the Mesha Inscription
andMoab (Atlanta: Scholars Press, 1989).
39. These radii appear at vastly different angles inside the circle among the attesta-
tions on the Mesha Stele.
40. For a treatment of this inscription, see A. Abou-Assaf, P. Bordreuil and A.R.
Millard,La statue de Tell Fekherye et son inscription bilingue assyro-arameenne
(Recherche sur les civilisations, 7; Paris: ADPF, 1982); S.A. Kaufman, 'Reflections on
the Assyrian-Aramaic Bilingual from Tell Fakhariyeh', Maarav 3 (1982), pp. 137-75;
J.C. Greenfield and A. Shaffer, 'Notes on the Akkadian-Aramaic Bilingual Statue from
Tell Fekherye', Iraq 65 (1983), pp. 109-16; A. Spycket, 'La statue bilingue de Tell
Fekherye', RA 79 (1985), pp. 67-68.
41. For the text of Azitwadda, see H. Donner and W. Rollig, Kanaandische und
Aramdische Inschriften (3 vols.; Wiesbaden: Otto Harrassowitz, 1964), 26. A facsimile
of Text A can be found in J. Naveh, Early History of the Alphabet: An Introduction to
the West Semitic Epigraphy andPalaeography (Jerusalem: Magnes Press, 1982), p. 56.
42. The Panammu I inscription is also known as the Hadad Inscription. For
treatment of this inscription, see J.C.L. Gibson, Textbook of Syrian Semitic Inscriptions
(3 vols.; Oxford: Clarendon Press, 1971), II, 13, PL 13.
43. See Donner and Rollig, Kanaandische und Aramdische Inschriften, I, 1; J.B.
Pritchard (ed.), The Ancient Near East in Pictures (Princeton, NJ: Princeton University
Press, 1954), pp. 456-59.
44. M. Lidzbarski, Handbuch derNordsemitischen EpigraphiknebstAusgewdhlten
Inschriften++++++++++++++++++++++++++++++
42 The Tel Dan Inscription
evidence of any strokes inside the rounded portions (i.e. the letter's inte-
rior). This means that if we are to understand this letter as a teth, the
interior stroke characteristic of a teth must be similar to the teth in Line 10
of the Mesha Stele (in the word mfrtf). This sees the interior stroke as
resembling two radii extending from the centre point out to the letter's
circumference on the top right quarter of the letter (similar to the hands of
a clock reading the time 2.15). This is indeed a very peculiar teth and one
that is rare and not consistent even within the Mesha Stele. The other pos-
sibility is to consider a shape similar to the teth in the Amman Citadel
inscription.
However, the second problem has to do with the placement of the
damaged letter on Line 2. If we follow the slant of the letters in Line 2 and
consider the damaged letter to be a teth, we find that it is placed well below
the normal writing base line for Line 2. Even if we allow for a change or
curvature of slant as we see in Line 3 directly below, we still find that a
teth would be very oddly placed. For this reason, and the one stated above,
we must have serious reservations about suggesting teth as the damaged
letter here.
We find similar problems with the suggestion of an 'ay in being the dam-
aged letter. Since both teth and 'ayin are of comparable height and shape in
other monumental lapidary inscriptions, the placement of the letter is trou-
blesome. Since we do have other attestations of 'ayin in Fragment A, we do
not need to appeal to other inscriptions for clarification of the letter shape.
It is quite clear that the 'ayin in Fragment A closely resembles a simple
closed circle. Although this accords well with the shape of the remnant of
the damaged letter, our main trouble is its placement well below the writing
base line. However, it is more plausible than a teth.
Another candidate for the damaged letter is lamed. We have numerous
examples within Fragment A with which to work. They present curvature
at the bottom left portion of the letter that fits well with the letter remnant
here in Line 2. In particular, the lameds of Lines 6-8 provide us with good
templates. The advantage of suggesting a lamed is that we have no prob-
lem with the letter's placement. Since lamed is a tall letter, the added
height fills the displacement with the writing base line (see Fig. 3.10). As
such, lamed must be considered a logical choice for reconstruction of this
damaged letter.
Schniedewind also reconstructs this letter as a lamed, but he does so on
the basis of'computer enhancement and imaging' and the assumption that
the letters reconstruct a name (as opposed to common noun, verb, and so
3. Epigraphical Analysis 43
45. Schniedewind, 'Tel Dan Stela', p. 77; Schniedewind and Zuckerman, 'A Possi-
ble Reconstruction', p. 89.
46. The biblical text portrays a certain Ben-Hadad (i.e. Bar Hadad) as Hazael's
predecessor. However, there is considerable doubt as to the reliability of this informa-
tion. This issue of the succession of Damascene kings will be dealt with in Chapter 7.
47. Schniedewind and Zuckerman, 'A Possible Reconstruction', pp. 90-91.
48. These issues will be discussed further in Chapter 7.
44 The Tel Dan Inscription
place, Schniedewind and Zuckerman would have realized that lamed is not
'the only letter that fits the spacing, orientation and line height'.49
Figure 3.10. Beginning of Line 2 (left) and possible restoration oflamedfor damaged
letter (right). The restored lamed is modelled on the lamed in Line 11 (detail of facsi-
mile drawn by Ada Yardeni).
Figure 3.11. The position of the deep scar (left) and its restoration together with the
damaged letter as a yodh (right) (detail of facsimile drawn by Ada Yardeni).
Figure 3.12. The remains of a stem belonging to a damaged letter at the beginning of
Line 2 (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z.
Radovan [detail]).
This, however, would see the stem as longer than the usual form attested
in Fragment A. The letter waw has similar problems in that we would need
to posit a longer stem than is usual. The size of the waw in Fragment A is
far from regular. A comparison of the waw at the beginning of Line 3 with
the waw in Line 5 demonstrated this. However, if we are to restore this
particular letter here in Line 2 as a waw, it would need to have the longest
stem of all waws attested in Fragment A.
A more plausible suggestion is the letter samekh. As in the case ofheh,
we would need to see the engraved point as part of the lowest horizontal
stroke. However, unlike the heh, the other attestation of samekh in Frag-
ment A indicates just how high on the stem this horizontal stroke actually
appears. As such, the samekh is a very suitable candidate for restoration.
Less can be said for the letter qoph. Although the engraved point that is
still visible to the left of the stem remnant is just one point, it would have
to be part of the round head of the letter qoph. If this was the case, we
might expect to find more of the head of this letter still extant on the
written surface because of the fact that it is round.51 However, the nature
of breakage and erosion along this portion of the stone means we cannot
maintain this expectation without doubt. Thus, it must not preclude us
from considering a qoph as an option for restoring the letter, especially
considering that the stem would be the right height.
We may also suggest the letter resh. Like the qoph and samekh, the stem
is of appropriate height. The main problem with resh, however, is seeing
the engraved point as the horizontal stroke of the letter's head. Examina-
tion of the other reshs on Fragment A reveals that this horizontal stroke
tends to weigh down on the right. As such, we would expect to be able to
see more of this stroke on the extant surface. Yet, as was the case with the
letter qoph, the damage inflicted on the stone at this point leaves enough
room for doubt, making resh another possible option.
In order to restore this figure as a sadhe, we would have to consider the
small engraved point to the left of the stem as irrelevant to the figure. Since
this engraved point is a deliberately carved stroke, we cannot regard it
simply as a scarification mark. Rather, it would need to be accounted for in
another waynamely, as part of another letter.
At this point, it must be noted that these two damaged letters before the
first word divider on Line 2, whatever the first of them is, appear to be
wider spaced than most other letters we observe on Fragment A. Through-
51. My previous suggestions have regarded the engraved point as part of a straight
line.
48 The Tel Dan Inscription
out the fragment we do find the tendency for two letters to be wider spaced
if the second letter is ayodh. For example, the first full word of Line 2,
"OK, has significantly more distance between the beth and yodh than
between the 'aleph and beth. The reason for this idiosyncracy is the tail of
the yodh which occupies a low part of the letter and which extends out to
the right. We find a similar phenomenon in the first word of Line 3, DZDCTl.
The first two letters of this word, waw and yodh, are more widely spaced
than the other letters because of the yodh's tail. This is certainly not a
universal peculiarity which shows itself in every instance of the letter yodh
on Fragment A. However, we may note that it does occur and why it
occurs. This would indeed lend more impetus to the suggestion that the
damaged letter immediately before the first word divider on Line 2 is in
fact ayodh.
However, we would still need to maintain an abnormally large space
between the restored yodh and the long-stemmed figure preceding it. To
whichever letter we restore the long-stemmed figure before the yodh, the
gap is still very large. We cannot propose a word divider in this gap because
we cannot supply an adequate meaning to a singular yodh as an independent
lexeme. Therefore, it is not a viable option to have such a large gap between
these letters.
Our only other alternative is to place another letter between the long-
stemmed figure and the yodh preceding the word divider. This means that
the engraved point can be accounted for if the long-stemmed figure is a
sadhe. This extra letter would by physical necessity have to be thin. It
must also have no long stem since no such stem is visible on the stone.
The only candidates for this letter are daleth, heth and 'ayin. The thinnest
letter in Fragment A, waw, is ruled out as a possibility because we can see
no traces of the stem. The very bottom part of the stem would still be visi-
ble on the stone. Therefore, we cannot consider it here.
One way of discovering what letter should be placed here is to test the
various possible combinations of letters. In doing this, we are testing two
particular points. First, the letters must fit physically beside each other
without any overlap. Second, the letters must make lexical and contextual
sense.
The only combinations of letters which fit into this space and align with
the carved strokes are "HD, HU, TID, TTH, *VD and TiJ. This means that the
long-stemmed letter can only be a samekh or a sadhe. Physically, the letter
following can be either of our three candidates (daleth, heth or 'ayin).
From these six combinations, three make lexical and contextual sense.
3. Epigraphical Analysis 49
These are:
1. '[l]D[n]: 'pious acts of or 'my piety'.
2. <?[n]iJ[]]: 'triumphs of or 'my triumph(s)'.
3. s[U]D[ft]: 'journeys of or 'myjourney(s)'.
With these possibilities, we must note that the second and third options
require the addition of another long-stemmed letter before the figure that
has left part of its stem on the fragment. Option 2 requires the addition of a
nun and option 3 requires a mem. The difficulty with both of these is that
the bottom portion of these letters' stems should still be visible on the
stone. The line of the stone's breakage is such that we would expect to see
part of the bottom of a long-stemmed figure if one was originally there.
Therefore, the second and third options are unlikely and I suggest recon-
structing the text to read in accordance with the first option, "'[!]&[PI].
Before this heth, there is room enough for two letters or, alternatively, one
letter and a word divider. Since no word dividers appear as the first char-
acter on any line of Fragment A, and since this does not occur in other
monumental lapidary inscriptions, we can plausibly say that a word di-
vider may be placed directly before the heth of ''[lib [PI]. Thus, a single
letter may have been the first character of Line 2, carried over from a word
at the end of Line 1. Equally plausible, however, is the prefixing of two
letters to the word ''[Tib[PI], such as a waw conjunction and a prefixed
preposition. This is the more plausible option contextually as will be seen
in our textual commentary on the text. We propose the prefixing of-31 to
* [~f ]D[PI ] in order to gain the reading * [~I]b[PQl ] ('and because of the pious
acts of or 'and because of my piety').
We move on, then, past the first word divider to a fully preserved
'alephthe first letter of the word "OK. The incisions of this 'aleph are
quite deep so that the walls of the incisions are quite sheer as opposed to
the round incisions seen previously (see Fig. 3.2). The vertical spine of the
'aleph appears to have been the final stroke of the letter etched into the
stone since it makes a decisive dissecting cut across the other two strokes,
rather than being dissected by these two strokes. The bottom horizontal
stroke also has a notable kink, close to the leftmost point of the letter.
Following the 'aleph comes the figure of a beth. The tail of this letter is
very curved and a chip has been sustained right at the point where the cur-
vature starts. Since it occurs at precisely this point, the chip may be indica-
tive of a finishing point which was hammered slightly too vigorously. Thus,
the tail of the beth was not carved as one stroke with the spine of the letter.
Rather, the figure was initially carved like a daleth with the tail being the
50 The Tel Dan Inscription
last stroke added. The evident direction of carving displayed by the strokes
suggests that this letter was also carved with the engraver situated along the
inscription's left edge. One curious point about this particular letter is that
the spine of the figure does not quite meet the diagonal stroke of the 'head'
at the topmost point. The outermost contours of the strokes meet, giving the
impression from a distance that the bottom of the incision channels also
meet. However, two deeply chiselled points are clearly visible at the top of
each respective stroke, showing that the bed of the incisions did not meet.
This highlights the fact that, at the time and place in which the inscription
was carved, the letter beth was clearly written with a pointed, rather than
rounded, headan important epigraphical datum.
The third extant letter on Line 2 is a yodh. It is quite difficult to ascer-
tain how this letter was carved, but it is clear that the top and bottom hori-
zontal strokes were not carved as one continuous stroke along with the
letter's vertical spine.
Following this yodh is another word divider that is not as round as those
seen previously. There is a clear deep point in the incision, indicative of
the single stroke of the engraver's chisel. This is telling for, since the word
divider is but one small chip, it highlights the direction of the engraver's
chisel. As an engraver hammers the final portion of a stroke into the stone,
an impression of the chisel's point is left at the deep point. In this case, the
impression is in the uppermost part of the word divider, indicating that the
chisel was aimed in an upward direction and angled slightly to the right.
This further strengthens the suggestion that a right-handed engraver was
positioned at the left edge of the inscription while carving.
Another yodh marks the beginning of a new word. This yod++++++
from the preceding one in that it appears to be more angular. Oddly, the
top horizontal stroke has not been carved smoothly and straightly. It
appears somewhat crooked, almost like a squat circumflex accent.
The next letter is a samekh. The stem of this letter seems to be carved
slightly deeper than the three horizontal strokes that it intersects, marking
it out as the final stroke carved in this letter. It is clear that this vertical
stem was also carved in an upward direction, due to a deep point at the
topmost extremity. This deep point has resulted in what looks like a bump
along the top horizontal stroke. This top horizontal line has evidently been
carved from left to right, while the deep point of the bottom horizontal line
indicates that it was carved from right to left. There is also some minor
erosion along the right half of this bottom horizontal stroke. It is, however,
quite difficult to determine the direction in which the middle stroke was
3. Epigraphical Analysis 51
carved. This fact may indicate that it was carved with much more care, and
so slightly shallower, than the other two horizontal strokes of the samekh.
We thus seem to have a scenario in which the engraver carved the top and
bottom strokes first, after which the middle stroke was engraved carefully
so as not to make any careless chips in the confined space which it occu-
pies between them.
What this samekh demonstrates, along with the letters already observed,
is that the engraver was restricted in carving vertical strokes upwards, but
that no similar restrictions governed the direction in which he carved hori-
zontal strokes. Thus, the suggestion that the engraver carved from one of
the flank edges of the inscription is confirmedthat is, the left edge assum-
ing a right-handed craftsman.
The line of breakage along the left side of the fragment has damaged
half of the next letter. What has remained is the right sector of a circular
stroke, with a stem extending below it. This is enough to identify the letter
as a qoph without any doubt. Some abrasions on the surface of the stone
have also been inflicted with the stone's breakage.
Therefore, we may decipher Line 2 as: [.. .]pD'''OK4[~T]b[rQ'l].
Line 3
The beginning of Line 3 has been preserved with the right edge of the
inscription intact. At this latitude of the stone, the right edge is severely
eroded, considerably more than it is at the level of the final line (Line 13)
of Fragment A. Some of this erosion has impinged upon the first letter, a
fully preserved waw, particularly over its stem. However, the erosion to
the letter strokes is still slight and does not hinder its identification. A deep
point is visible at the top extremity of the stem of this waw, indicating the
upward direction of vertical strokes we have come to expect. Furthermore,
two deep points along the 'hook' of the wawone at the apex and another
at the topmost pointare consistent with our theory of the engraver's
position during carving. It also suggests that the stem was the first stroke
of the letter carved into the stone since the 'hook' has been carved away
from it. One further aspect of this letter is that it is spaced slightly further
away from the next letter than is usual. This is probably due to the fact that
the next letter is ayodh and the scribe who chalked the text onto the stone
appears to have distanced his yodhs from the preceding letter in order to
accommodate the tail.
After the fully preserved and quite angularyodh, we see a clearly legible
sin. Directly above the midpoint of this sin is a pock mark that appears
52 The Tel Dan Inscription
this heh we have a very deeply carved kaph. The left 'finger' of the letter
has some slight chipping at its extremity. All the 'fingers' however do
meet at one point on the stem. A very deep word divider comes after this,
which confirms our theory of where the engraver was positioned. The en-
graver's tool has left its mark quite well in the single chip needed to carve
the word divider. It demonstrates that the nib of the tool was shaped
exactly like a flat-bladed screwdriver. Yet, in addition, this mark of the
engraver's tool is aimed upwards to the right, demonstrating the angle at
which the tool was struck into the stone. This is the most natural angle at
which to aim the tool if one is situated along the left edge of the inscrip-
tion. It would be an unnatural angle from anywhere else.
Following this word divider, only two letters are still extant. The first is
an 'aleph that shows some signs of erosion along the top right diagonal.
Again, the vertical stroke forming the spine shows evidence of having been
carved from bottom to top. This stroke is also quite shallow. The second
letter is a lamed carved on a notable right slant. This could perhaps be
accounted for by the scribe's outstretched posture when chalking the text
onto the stone. The stone is slightly damaged in the middle of the letter,
almost lending it an appearance as a small Latin 'b'. The line of breakage
of Fragment A impedes on this lamed at the point where the 'shank'
begins to hook right. This being the case, we cannot see any shallow point
where we expect the two strokes of this letter to have started. However, we
may reasonably expect the 'shank' of the lamed to have been carved from
bottom to top, after which the hook of the letter was carved. Since the
extremities of these two strokes are visibly deep, we would be right in
surmising this.
Thus, we can decipher the characters on Line 3 as: ~[!T''3KO;DEn
[...]^.
Line 4
Line 4 begins with a neatly carved resh. The stone is slightly eroded here
because of the belt of erosion that encompasses the entire right edge of the
fragment. As with the 'aleph that follows, the vertical stroke again has
been carved from bottom to top. The spine of the following 'aleph is also
quite shallow. Very close to the leftmost point of the 'aleph is another
lamed. This lamed stands more upright than the one at the end of Line 3,
and we can see that the strokes were indeed carved outward from where
the 'shank' bends into the hook. Another word divider follows this letter
which again clearly shows the direction and position in which the engraver
carved his strokes.
54 The Tel Dan Inscription
52. A partially preserved qoph was observed at the very end of Line 2.
3. Epigraphical Analysis 55
Presumably this was caused during the inscription's breakage, as was the
erosion which impinges on the preceding word divider. The enclosed tri-
angular area of the 'aleph has been chipped completely out in a way simi-
lar to that of the first 'aleph in Line 3. However, the leftmost portion of
this 'aleph has been totally damaged because of a very deep lacuna in the
stone. There is no difficulty, though, in identifying this letter as an 'aleph.
This lacuna presents an interesting problem that received only cursory
attention from Biran and Naveh, and none from any other scholar con-
cerned with the fragment. The problem is the size of the lacuna and the
gap between the 'aleph just mentioned and the beth which emerges clearly
(albeit slightly damaged) at the end of the lacuna. Biran and Naveh suf-
ficed to say that this gap did not warrant the restoration of a letter in
between the 'aleph and the beth. Rather, they suggested that a small
blemish already existed on the stone prior to the engraving of the inscrip-
tion and it was this which accounted for the larger than usual gap. This
blemish, they claim, then became larger over time.53
However, Biran and Naveh have denied two crucial pieces of evidence
in saying this. First, the lacuna is quite deep; it is not superficial. Only a
significant force concentrated onto this small portion of the stone could
have made the lacuna. Its depth suggests that it had not been caused by an
originally smaller chip that had been unwittingly enlarged over time. Had
this been the case, we would expect the edges of the lacuna to be quite
shallow with probable patches of erosion blurring the edges. This is cer-
tainly not the case. In fact, one could almost mistake the lacuna as the
result of deliberate chiselling. The edges are not shallow; they are quite
deep. Also, there is no blurring around its edges to suggest that it is the
result of natural erosion. As such, there is no reason why the 'aleph and
the beth should be so far apart.
In fairness to Biran and Naveh, this could simply be an accidental wide
spacing of the letters. However, the second piece of evidence dismisses
this possibility. Protruding from the top of the lacuna, midway between the
'aleph and the beth, is the remnant of a definite carved stroke. This stroke
extends upwards and to the right, almost touching the bottom of the stem
belonging to the kaph in the line above (see Fig. 3.13). It comes as no
surprise that this stroke was not observed by any other epigrapher because
the photograph of Fragment A published by Biran and Naveh is mislead-
ing. The lighting in this photograph, taken by Radovan,54 gives the impres-
sion that two scars protrude from the lacuna. This is not quite the case with
the real fragment. There is some slight scarification leading into the lacuna,
but this is very shallow, unlike the impression given by the Radovan pho-
tograph. The 'scar' to the right of this, however, is a deliberately carved
stroke and we must conclude that it is the remnant of an extra letter that
has hitherto gone unnoticed by the scholarly community. A more con-
trasted image of the fragment shows the depths of the scar and deliberately
carved stroke a little more clearly (see Fig. 3.13).
Figure 3.13. Detail from the Radovan photograph (top) of the lacuna on Fragment A,
Line 4, in which the scarification appears overly accentuated. A facsimile (bottom)
labels the lacuna, scarification and the deliberately carved stroke representing the
remains of a letter (Tel Dan Excavations, Hebrew Union College, Jerusalem; photo-
graph: Z. Radovan [detail]; facsimile drawn by Ada Yardeni [detail]).
Having observed this carved stroke, we must now identify it with a let-
ter for restoration. The slant and positioning of the letter leave only one
possible candidate: a lamed. The only other letter that could possibly fit is
a taw, but it has numerous problems. First, no other taw in either Fragment
A or Fragment B sits quite so high. Even allowing a concession for this,
3. Epigraphical Analysis 57
Figure 3.14. An electronically altered image (left) showing the possible original
appearance of the damaged letter and a facsimile (right) of the suggested restoration
(Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan
[detail]; facsimile drawn by Ada Yardeni [detail]).
Line 5
The fifth line of Fragment A begins with a smoothly eroded 'aleph. This
erosion is part of the belt of erosion that covers the right edge of the entire
fragment. Following the 'aleph, we come to our first attestation of the
letter nun. Because of the erosion along the inscription's edge, this letter is
also quite shallow. Nevertheless, we can make out the deep and shallow
points at the extremities of the stem. Once again, these indicate that the
letter was carved from bottom to top. Similarly, the vertical hook of the
stroke was carved in an upward direction. The direction of the hook's
horizontal stroke is harder to discern.
The nun is followed by a heh, the stem of which is fairly short.57 As we
have seen with other occurrences+++++++++++++++++++++++++++++
the letter to be carved, and it was carved upwards. A word divider follows
very closely, almost touching the top 'bristle' of the heh. It is quite small
and shallow due to the smooth belt of erosion.
After this comes a waw that is also quite shallow due to erosion. How-
ever, this letter shows quite distinctly the starting and finishing points of
both the stem and the vertical stroke of the hook, again confirming the
upwards direction of carving. This is followed by an odd looking yodh;
odd because of its cramped appearance. The tail of the letter is relatively
shorter than those of other yodhs in the fragment, but it also bends up-
wards quite significantly. After this is a heh which is also shallow. Yet, in
keeping with previous letters, the stem was inscribed upwards and was
evidently the final stroke carved. The bottom two 'bristles' of the letter
barely join the stem. A kaph follows this heh, which sees the bottom part of
the stem damaged because of a pock mark in the stone. Yet, since we have
the top of the letter, we can see deep points at the top extremity of every
stroke. This suggests, once again, that the vertical strokes were carved up-
wards and the horizontal 'finger' was carved leftwards. The top of the stem
also comes very close to touching the top 'bristle' of the preceding heh.
A deep word divider comes next which is unaffected by the pock mark
damaging part of the kaph. As was the case in a previous word divider, we
can observe the definite shape of the nib of the engraver's tool in the upper
right part of the indentation. Again, the engraver was evidently situated at
the left flank of the inscription.
Another heh follows this word divider. The very bottom of the stem has
been damaged by the pock mark which intrudes over the stem of the kaph.
This heh follows the exact same pattern of carving as all the preceding
hehs: an upward direction with the stem carved last.
Two daleths are seen after this heh. The first daleth has been slightly
damaged along the bottom horizontal stroke. This is due to a small but
quite deep pock mark on the surface of the stone which must have been
57. The Radovan photograph of Fragment A gives the false impression of a long
stem.
3. Epigraphical Analysis 59
Line 6
The first letter of Line 6 is ayodh that has been smoothly eroded. It bears a
great resemblance to the yodh mentioned above at the end of Line 5. After
this is an extremely shallow word divider. It appears that the engraver did
not inscribe this word divider very deeply to begin with and the effects of
60 The Tel Dan Inscription
erosion have weathered the depth of the engraving down even further. So
shallow is it, in fact, that were it not placed in a gap between the preceding
yodh and the following letter, we could surmise that it was an accidental
blemish on the stone.
The letter that follows is another mem, made shallow because of erosion
along this flank of the inscription. In turn, this is closely followed by a very
neat lamed which appears almost pressed up against the mem. A shallow
kaph follows next which again displays evidence of having been carved
upwards, with the horizontal 'finger' being carved outwards from right to
left. A yodh follows this, which is of the more angular form, followed by
another word divider showing clear evidence of the engraver's position
during carving.
A waw is seen next which has undergone some erosion and very slight
chipping along the carved strokes. The deep points at the top extremities
of the vertical strokes again testify to the letter being carved upwards. An
'aleph follows after this, which is very regular and displays all the normal
characteristics of previous 'alephs. There has been some slight abrasion of
the stone across the bottom of the letter.
Next to this we find a qoph similar to previous examples. The letter is
carved upwards and the head was started from the bottom left corner. This
is followed by a taw which has been carved so well that it is impossible to
tell whether the upright or cross-bar of the letter was carved first. What is
discernible, though, is that the upright was carved from the bottom upwards.
To the immediate top left of this taw begins a pock mark which spreads
across to the 'shank' of a following lamed. Inside the pock mark we can
observe a slightly hollow pocket within the actual stone showing that the
surface of the stone here collapsed under great impact to form the pock
mark now visible. That is, this was a natural weak point in the stone due to a
hollow bubble. Despite this damage, the lamed is easily discernible.
Following this is a word divider that shows the angle at which the
engraver hammered the stroke quite distinctly. This angle, however, is
slightly more upright than most others before or below it. Nevertheless, it
is clear that the engraver was still situated along the left flank of the
inscription when he carved this character.
After the word divider, we have the remains of what is clearly a mem.
The bottom of the letter has been damaged and destroyed by the semi-
circular recess which we previously suggested was caused by a hammer
used to deface and destroy the original inscription. The remaining portion
of the stem on this letter precludes it from being a sin.
3. Epigraphical Analysis 61
The recess, however, has almost completely obliterated the next letter.
The only portion of this letter still visible is a stroke protruding up from
the recess and leaning to the right. In their publication of Fragment A,
Biran and Naveh proposed restoring this as a nun.58 This seems a plausible
suggestion at first, but there are certain difficulties with it. First, it would
require us to see this stroke as the upper left portion of a nun. The forms of
nun attested in Fragment A (and Fragment B) show that having this upper
left stroke on such a decided slant is irregular. We do find such a form in
Line 11 of Fragment A but it is unique among the attested forms. How-
ever, we cannot dismiss the possibility of a nun on that basis alone. Yet, a
second objection appealing to the position of this visible stroke must be
raised. If this letter was actually a nun, we must say that it is abnormally
close to the preceding mem. In fact, when we superimpose the images of
other nuns from both Fragment A and Fragment B over this position, we
find that they either overlap with the preceding mem, or are placed so close
as to make the mem and nun look connected. It is as though they were
strokes of one wide and incomprehensible letter. Furthermore, the visible
stroke above the recess is carved in a very high position for a nun. It is to
be noted that it sits higher than the topmost point of the mem before it.
Again, this would be an irregular placement for a nun. Therefore, if we are
to maintain that this letter could be restored as a nun, we must say that it is
irregularly thin and irregularly high. The number of objections that may be
raised against this, however, means that restoration of a nun is theoreti-
cally possible but highly unlikely.
We did, however, observe a similar stroke protruding from the top of the
lacuna in Line 4 where it was concluded that a lamed needed to be restored.
Its placement so close to the preceding 'aleph in that line posed no serious
problem for restoring a lamed because of the nature of the lamed's shape.
We inevitably must come to the same conclusion here since both remnants
are for all intents and purposes identical. Thus, we should opt for restoring
a lamed'here as the most plausible candidate for this damaged letter. In their
second publication which annnounced the discovery of the B fragments,
Biran and Naveh changed their opinion to also restore this letter as a
lamed.59 However, this was evidently not based on comparison with the
stroke protruding above the lacuna in Line 4. Rather, it was based on their
arrangement of the fragments and the subsequent reading of the text that
they offered.
Figure 3.15. Detail of thefacsimile drawn by Yardeni (left) showing the damaged letter
at the edge of Line 6, Fragment A, and an electronically altered image of the same
area (right) showing a more accurate representation of the damaged letter.
We cannot find any other letter fitting this shape, besides waw. Yardeni
suggested that this was the tip of a nun.63 However, this has only taken
into account the rightmost of the three strokesa misperception that
obviously influenced the drawing of the facsimile. The letter nun also does
not attest to a form with the main stem extending above the 'hook'.
60. This does not preclude a lamed occurring elsewhere along this line in the
original inscriptiononly on the rest of the line as it appears on Fragment A.
61. Biran and Naveh, 'An Aramaic Stele Fragment', p. 88, Fig. 7.
62. Biran and Naveh, 'An Aramaic Stele Fragment', p. 89, Fig. 8. In this facsimile,
the edge of the recess covers more of the strokes than actually is the case and only the
rightmost of the three strokes has been drawn.
63. Biran and Naveh, 'The Tel Dan Inscription', p. 16.
3. Epigraphical Analysis 63
Dijkstra, who also noticed only one stroke, suggested that it was the rem-
nant of a sin. He used images of the fragment to transpose letters over this
stroke and concluded that 'the small stroke almost exactly coincided with
the rightside of the letter, sin ofysr+++++++++++++++++++++
case and neither does it take into account the exact nature of the strokes. In
his extravagant reconstruction of the text of Fragment A, Puech restored
this letter as an 'aleph. In the facsimile published with his article, we ob-
serve that Puech noticed more than one stroke at the edge of the fragment
here.65 However, he misconstrued the precise nature and orientation of
these strokes, and so came to a faulty conclusion.
The only letter we may restore here is a waw. Indeed, we find almost
identical forms of waw elsewhere among the fragments. The waw occur-
ring earlier in Line 6 shows a form where the stem displays a notable kink
at the junction with the 'hook'. The letter remnant at the edge of the frag-
ment displays this exact same feature. The second waw in Line 9 also
shows similar characteristics. Even closer is the form of waw in Line 3 of
Fragment B. This waw not only shows a kink in the stem, but a much more
open 'hook', also. This compares almost exactly with the remnant we find
at the edge of Line 6 in Fragment A. As such, we must restore this letter as
a waw.
This has implications for restoring the letters that would originally have
been carved between this restored waw and the lamed that we also restored
a little earlier. There is room enough for two very closely compacted let-
ters or for one uncramped letter and a word divider. The contours of the
recess at the edge of Fragment A show just enough of the written surface
to suggest that the latter is the case. If there were two letters in this gap,
they would be so close to each other and the enclosing letters that we
would expect to see remnants of at least the second letter preceding the
final waw. This would be the case even if one or both of the letters were
thin in shape, like a waw or nun. However, even allowing a concession for
this, a waw or nun would pose considerable philological problems for
reading the text. Yet, we must allow for the possibility of two very small
and closely compacted letters. On the other hand, having one letter and a
64. M. Dijkstra, 'An Epigraphic and Historical Note on the Stela of Tel Dan', BN
74 (1994), pp. 10-14(11).
65. E. Puech, 'La Stele Arameene de Dan: Bar Hadad II et la coalition des Omrides
et de la maison de David', RB 101-102 (1994), pp. 215-41 (218-19, 224-25); see
especially p. 219, Fig. 1. Puech also considered the possibilities of a lamed, yodh or
kaph, but dropped these in favour of 'aleph.
64 The Tel Dan Inscription
Line 7
Due to the gradual erosion on the right side of the stone, the strokes of the
first few letters on Line 7 are found to be shallow. The first letter is a
regular kaph, followed by a very neat beth and a word divider that shows
evidence of the engraver's position along the left edge. A waw follows
this, which has been eroded even more than the other letters around it.
This lends the contours of the strokes a very soft edge. In contrast, an
'aleph follows this waw which has significantly sharper edges and con-
tours due to less erosion across the surface of the stone. Next to this we
encounter a rather angular lamed in which the curves of the expected
shape have been sharpened almost to corners.
Our first attestation of the letter peh follows this. It is simply in the shape
of an upside-down lamed. In contrast to the shape of the preceding lamed,
this peh is well curved but is fairly shallow because of some erosion. A
deep point in the incision is observable at the topmost point of the letter
where it begins to curve downwards. This suggests that the letter had been
carved in two stages. The first stroke appears to have been the short curve
at the top of the letter, after which the hanging stem was carved from the
bottom up. Thus, the two strokes met at the one point, resulting in one
common deep point.
A nicely curved yodh comes after this peh. The tail of this yodh almost
imposes upon the space of the preceding peh. We should also note that its
uniform curvature provides a good template for thQyodh restored immedi-
ately before the first word divider on Line 2.
After another word divider, we encounter our second example of the
letter peh. It is very similar to the peh encountered two letters before, the
only difference being that the curve appears slightly more open. It has been
skilfully carved to give the impression of one fluid stroke. However, as
with the previous peh, a deep point is apparent at the top of the letter indi-
cating it was carved with two strokes meeting at one point.
Following closely after the peh is a regular resh. The only noteworthy
point of this resh is that the stem ends very close to the bottom of the peh.
3. Epigraphical Analysis 65
A regular sin follows this, after which we observe a fairly deep word
divider. At this point, the line has been broken by the large recess. There is,
however, a minute trace of the next letter. It is a very small portion of a
stroke that is situated along the actual edge of the recess at the same latitude
as the word divider. Since it is such a tiny sample, it is impossible to sug-
gest any restoration for this letter based solely on the physical evidence.
We can, therefore, decipher Line 7 as: [...]^ns''SLWin.
Line 8
Line 8 begins with a shallow mem that has an odd slant when compared to
other mems in the fragment. This is followed by an angular lamed which is
similar to the lamed in Line 7, though rotated significantly in a clockwise
direction. The third letter of the line is a very neat kaph that displays all
the regular features of previous examples.
After this, we encounter a word divider notable for its placement. It
occurs higher up than most other word dividers, but it also sits directly
above the tail of theyodh that follows. That is, it imposes upon the space
of theyodh. In fact, if the word divider was to be removed, we could read
the first two words of Line 8 as if they were one word, for there is no
noticeable gap in which the word divider is placed. Nevertheless, the word
divider is a deliberately carved stroke. We cannot suppose that it is a chip
in the surface of the stone. It must, then, be registered as an inscribed
character of the text.
The yodh that follows is actually quite small. It is both thin and short
when compared with other attestations of the letter in Fragment A. The top
horizontal of the letter has also been eroded slightly. It is the first letter of
the word ^fcdEF, of which all the letters appear regular.
After the lamed of ^N"IET comes a word divider that has been made
shallow by a small patch of erosion. This eroded area extends across the
top of the next letter, a wow. After this waw we have half of a qoph. The
edge of the large recess follows precisely down the length of the qoph"?,
stem so that only the right half of the letter is still preserved. However, at
this point we encounter the micro-fragment that sits inside the large recess.
We can actually still observe the point from which the head of the qoph
was carved on this micro-fragment. This point is represented by a small
mark on the very edge of the micro-fragment in the same relative location
we have seen it in other examples oi'qoph.
After this qoph, we find the lower portions of two letters that are
preserved on the micro-fragment. The first is obviously a taw and the
66 The Tel Dan Inscription
second is a curved stroke that can only be the bottom of a lamed. To the
lower left of this lamed we do see signs of some abrasion on the stone
surface which, in the Radovan photograph, could be misinterpreted as a
word divider.66 However, the lighting in this photograph has simply
exaggerated the depth of the abrasion. At this point, the line breaks off.
We can decipher Line 8, therefore, as: [...]^flp>L'^~l^~[bQ.
Line 9
The first letter of Line 9 is a kaph that shows signs of wear due to con-
tinued erosion. The 'fingers' of the kaph show the most erosion. Inter-
estingly, the 'fingers' of this kaph do not meet uniformly at one point. It is
the last letter of a word which originally began on the previous line, the
end of which is no longer extant.
After a shallow word divider, we encounter a beth which is smaller in
size than other beths in Fragment A. The strokes which form the 'nose' of
the beth do not actually meet, though the wider contours of the incisions
appear to do so from a distance. After this we find a yodh which is quite
jagged in shape. This is followed by a taw situated under a small patch of
erosion that affected the yodh directly above it in Line 8.
Following this we have a daleth with a very short tail, a regular waw and
another daleth with a more regular tail. There is a slight gap between the
waw and the second daleth, which comes very close to the word divider
that follows. After this word divider we find a waw. The 'hook' of this waw
appears at an odd slant, lending it a 'heavy' appearance. An 'aleph follows
this, which possesses a rather long stem.
After this 'aleph, we have only the right portion of a letter damaged by
the large recess. It consists of an angled stroke that is consistent with the
rightmost stroke of a sin. A very close examination of the point of break-
age confirms this, for we find the slightest trace of a stroke angling up
from the bottom of the stroke still extant. We could argue for the restora-
tion of ayodh, but this would require us to see the visible stroke as the tail
of the yodh. Yet, none of the yodhs in the fragments displays such an
upwardly angled tail. The only yodh that comes close is the first yodh of
Line 3, Fragment B. However, each yodh attested in the fragments has a
spine which either stands upright or leans to the right. The slight trace of a
stroke at the point of breakage here shows that this stroke leaned to the
left. As such, we must posit a highly irregular form of yodh if we are to
66. Biran and Naveh, 'An Aramaic Stele Fragment', p. 88, Fig. 7.
3. Epigraphical Analysis 67
consider one at all. A sin, however, has no such problems and is therefore
to be preferred.
Beyond this damaged letter we have the remains of another letter on the
micro-fragment inside the large recess. The remnants of this letter consist
of a jagged shape equating with either a mem or a sin. Since we do not
have the rightmost point of this letter, we cannot appeal to the presence or
absence of a stem to help us distinguish between a mem or sin. However,
the preceding comparison of these two letters in Line 1 showed that the
leftmost stroke of all sins always leans to the left. In contrast, the leftmost
stroke of all mems leans to the right. With this knowledge we can safely
say that this damaged letter is a mem because the leftmost stroke leans to
the right. We cannot argue for an irregularly shaped sin here because the
evidence from the available sample of fragments is unanimous in this
difference between sin and mem. After this restored mem, we see a word
divider. At this point, the line breaks off.
We can decipher Line 9, therefore, as: [...]b&K>TnrV!>~[.
Line 10
The beginning of Line 10 has a shallow yodh, followed by a taw and a
very eroded word divider. This is followed by an 'aleph, the top of which
is also eroded. After this we see a resh. The stem does not actually meet
the diagonal stroke at the top of this resh, though from a distance one
gains the impression that the strokes do meet. There is also some erosion
at the 'nose' of this resh, affecting the following letter, a qoph. Interest-
ingly, the head of this qoph has a double crown unlike other attestations.
Another shallow word divider appears after this.
Next we find a heh and a mem, followed by yet another shallow word
divider. The last legible letter on the line comes after this in the form of a
lamed. There is some gap after the lamed before the line breaks off. Unfor-
tunately, the line of breakage occurs before the next letter.
Line 10, then, should be deciphered as: [...] L >Dnp"IKrr.
Line 11
A shallow 'aleph begins Line 11. After it we encounter our first instance
of the letter heth. The bottom 'rung' of the heth has been damaged by
erosion, but it otherwise appears to have been carved very neatly. The
vertical 'runners' of the letter have been carved from the bottom up while
the top two 'rungs' have been carved from left to right. Due to erosion, it
is impossible to tell in which direction the bottom 'rung' was carved. We
68 The Tel Dan Inscription
may notice, though, that the middle 'rung' is shallow in comparison with
the other strokes, suggesting that it was the last 'rung' carved with extra
caution.
After this heth we see a regular resh, followed by a wide, irregular nun.
The irregularity is that the top left stroke is on a considerable slant. After
another shallow word divider, we see a waw. The 'hook' of this waw is
fairly round rather than angular. The lamed that follows, though, is actu-
ally fairly angular. After it, there is a considerable gap to the next letter, a
heh. Oddly, the bottom of the stem bends to the left. The line breaks off at
this point, leaving only the tiny trace of another letter on the edge at the
same approximate latitude as the top of the heh. The portion must be the
very top right part of another letter. The nature of the incisions allows us
to narrow the candidates for this letter down to three letters: nun,peh and
qoph. The letter qoph is the least likely of these three, but the visible inci-
sions do not permit us to identify the letter more accurately than this.
We can decipher Line 11, then, as: [...?]nb>pnN.
Line 12
Very close to the edge on Line 12 we find a small indentation which could
pass as a word divider. It is much closer to the edge than any other inci-
sion along this right flank of the inscription. There are also numerous other
chips and abrasions in this corner of the fragment. Since we also have no
other evidence of a word divider beginning a line, we are justified in
regarding it simply as a blemish on the stone.
The first letter on Line 12 is a very angular lamed. There is considerable
erosion over this entire corner of the fragment, so the lamed is fairly
shallow. The kaph that follows shows significant signs of wear, particu-
larly around the 'fingers' of the letter. The nature of this damage suggests
that it had been sustained during breakage of the inscription as we also see
similar chipping immediately below the kaph on the next line.
After a very shallow word divider, we encounter a regular 'ayin. After
this we see another lamed which sits a good deal higher than the other
lamed at the beginning of the line. Like that lamed, it is of the angular
variety rather than the rounded variety. Continued erosion has worn away
the following word divider. A yodh appears after this, displaying curved
features rather than angular features. There are also some abrasions and
chipping surrounding this yodh. Last on the line, we find the greater part of
a sin. Only the bottom left portion of the sin has been broken off so its
identification is not in doubt.
3. Epigraphical Analysis 69
We may decipher the extant characters on Line 12, therefore, as: ~[L>
[.. .]ET v?^. Reconstruction of some of the non-extant letters will be left for
the textual analysis.
Line 13
The last line of Fragment A is quite short and is totally enveloped in the
belt of erosion along the right flank of the fragment. The line begins with a
mem of which the greater portion of the stem has been destroyed by break-
age of the stone. Immediately to the left of this is an area of concentrated
abrasion. This impinges on the next letter, which should be identified as a
sadhe. The stem of this sadhe is almost entirely visible and the first two
strokes off the stem are still visible. The rest, however, has been damaged
by the abrasions to the stone. As such, we cannot make out with any cer-
tainty the right-hand side of the letter. Its identification as a sadhe, though,
is not in doubt since no other letter in Fragment A compares with it, and the
sadhe of other inscriptions is comparable with what is still visible.
The next letter is a very upright resh. Between it and the preceding
sadhe, we find miniscule chips that must have been sustained during
breakage of the inscription. The entire stem is still extant. After it, we have
a considerably eroded word divider, followed by a regular 'ayin. Some
erosion has affected the top of this letter. The last letter extant on this line
is a lamed. Only the bottom left-hand portion of this lamed has been
broken off; it is still easy to identify. At this point, the fragment breaks off
completely. Some of the written surface below Line 13 is still extant, but
the fragment breaks off before the next line.
Line 13 can, therefore, be deciphered as: [...j^UOHQ.
Preserved Side
Some of the right side of the original inscription has been preserved on
Fragment A. The actual edge has been considerably weathered. The
preserved side is very smooth but shows numerous abrasions which should
be attributed to the inscription's breakage. There are, however, also signs
of smooth erosion similar in nature to the belt of erosion along the right
flank of the written surface. This suggests that the side of the inscription
was exposed to the elements in a similar way to the written surface. This
could have been sustained either during the time the inscription was on
display, or during its time as building material in a wall on the edge of the
piazza at the outer Israelite gate of Tel Dan. Since the weathering is con-
centrated near the edge, as it is in the written surface, it was most probably
70 The Tel Dan Inscription
67. Unfortunately, Cryer never published the evidence for these claims.
68. See Bibliography.
69. This was communicated personally to me.
3. Epigraphical Analysis 71
probably that of another rock. Thus, we expect them to have been sus-
tained when the stone was destroyed or, more likely, when it was reposi-
tioned in the wall at the perimeter of the piazza.
In addition to the physical evidence, it is hard to imagine why a poten-
tial forger would deliberately carve a stroke into the broken surface. It
would require the engraver to have been positioned behind the fragment or
for the fragment to have been lying face down on the ground. It is cate-
gorically impossible for such a stroke to have been inflicted accidentally
while carving the text on the written surface. We are, therefore, hard
pressed to find any logical reason to account for how a chisel mark came
to be on an exposed surface at any time.
Furthermore, we have to contend with the pattern of smooth erosion on
the right side of the written surface as well as the preserved side that meets
it. This weathering can only have been sustained over a long period of
exposure to the elements. Since Fragment A was recycled in the base of a
wall, it was located at a runoff point for rainwater. We may deduce that
this is how the belt of erosion along the right edge was made. That this is a
slow process testifies to the age of the fragment. It would have to have
been in this position in the wall at the perimeter of the piazza for many
years. Yet, the piazza at the outer gate of Tel Dan was only fully exposed
by archaeologists in 1992 and 1993. Fragment A was discovered in July of
1993. If Fragment A is a forgery, the forger must have weathered the stone
over many years for it would be impossible to accomplish this type of ero-
sion by artificial means. Such a situation is too incredulous to be legitimate.
The only reasonable explanation for the weathering on the stone is that it
was exposed to elements in antiquity prior to the piazza being buried by
debris and subsequent strata. We can therefore be assured of the authen-
ticity of Fragment A.
Cryer's claims of forgery also call into question the integrity of the
excavation team at Tel Dan, led by Biran. If the fragment was a forgery,
we must posit one of two theories. Either the forgery was the work of the
excavation team itself who have since falsified information in order to
uphold an image of authenticity, or the fragment was planted at Tel Dan
by an individual (or individuals) with exceptional epigraphical, palaeog-
raphical, philological and historical expertise.
Neither of these positions can be maintained with any sense of sobriety.
Not only would it require someone of such skill and expertise to forge an
inscription such as Fragment A (and indeed Fragment B), they must also
have had easy access to the find site at the entrance to the Israelite outer
72 The Tel Dan Inscription
gate at Tel Dan. Thus, if Fragment A were a forgery, it would have to have
been planted in 1992 at the earliest. Fragment A's position of discovery in
the base of a wall also means that anyone who planted the fragment would
have had to undertake extra excavation in order to remove some stones,
replace them with the fragment and then restore the area to make it look
untouched. Thus, in addition to being an inscriptions expert of the highest
calibre, the individual must also have been in league with, or actually been,
a skilled archaeologist. Such a person (or persons) are likely to have been
a senior member of the excavation team at Tel Dan. That is, we expect the
individual to have been part of the active scholarly community. However,
the integrity of Avraham Biran as an archaeologist cannot be questioned.
His many years of research and excavation into the archaeology of the
Levant stand him in good stead as an archaeologist with absolute integrity.
Similarly, we can label none of his excavation team during 1992-93 with
such unscrupulous behaviour.
Cryer advocated that all three fragments must be seen as part of the
same forgery. Indeed, if Fragments B1 and B2 represent a separate inscrip-
tion, Cryer's accusations fall through. Yet, Fragments Bl and B2 were
both discovered a year after Fragment A in 1994. It was not until that year
that the find sites of both these fragments were exposed. If Fragments Bl
and B2 are also forgeries, we must suggest that they were planted in the
year of their discovery (1994) after the exposure of their find sites and
after the discovery of Fragment A. Such inconsistency does not allow us to
maintain any of the three fragments as forgeries. In short, the accusation of
forgery against the Tel Dan fragments is groundless and should be sum-
marily dismissed.
Fragment A: Assessment
Fragment A is shown to be in extremely good condition. The signs of wear
on the written surface are mostly to be attributed to the breakage of the
stone and the subsequent transference of the fragment to a new location
for use as building material. The consistent smooth erosion along the
entire right edge of the fragment, however, can only be attributed to weath-
ering during antiquity. Since this weathering is confined to one edge of the
fragment, this edge must have been exposed to the elements more than any
other area on Fragment A. Furthermore, since the weathering can be seen
on the side of the fragment, we cannot maintain that weathering occurred
while the inscription was part of an ashlar block in a wall. This erosion
3. Epigraphical Analysis 13
must have occurred when the fragment was in secondary usage in the base
of the wall at the perimeter of the piazza.
We may compare the state of the written surface with those of other
monumental lapidary inscriptions. A picture then emerges of just how well
preserved the written surface of Fragment A is. Other inscriptions, such as
the Melqart Stele, the Mesha Stele and the Zakkur Stele, all show a greater
degree of wear.70 This leads us to either of two conclusions. First, the in-
scription which Fragment A represents may have been shielded from the
elements in its original display position. Our second possibility is that the
original inscription was not on display for very long at all before its
destruction and reuse as building material, and that it was in its secondary
stage of use that most of the weathering occurred.
From the large semi-circular recess along the left edge of Fragment A,
we may glean that the original inscription was a free-standing stone. Since
this recess is the result of a deliberate hammer blow, the original inscrip-
tion must have been able to topple or be easily damaged by such a hammer
stroke. Thus, the inscription's nature as a stele is confirmed.
70. Half of the Mesha Stele on display at the Louvre, Paris, is stone replica. This is
because much of the original stone inscription was destroyed in the nineteenth century.
The original parts of the inscription still in existence today show signs of considerable
wear.
71. This was stated by Avraham Biran in personal correspondence with me. Special
thanks go to Biran for his assistance here.
74 The Tel Dan Inscription
When this datum is coupled with the scribe's tendency to lift the lines,
we have a picture of the scribe nearing the end of the text. This is also
confirmed by the fact that many of the letters almost encroach on the
letters of the next line as though there was only a short space still available
on the written surface and the scribe needed to put as much writing as
possible in the last few lines. This is most noticeable in Lines 4 and 5 of
Fragment B.
The tendency to lift the slope of the lines is consistent with a right-
handed person writing while positioned slightly to the left of the centre
line of a writing surface. This is the most convenient position to take when
writing from right to left in order to avoid crossing the right hand and arm
across the body too far. This is especially desirable when the written
surface is fairly close to the body as Fragment B appears to have been.
Thus, we have a situation that sees Fragment B as the lower portion of
an inscription, though not the very bottom portion. This is confirmed by
the neatness of the letters, the encroaching of some letters onto the line
below, and the tendency to lift the slope of the lines.
78 The Tel Dan Inscription
Figure 3.18. Diagram of the angle between the written surface and the newly
exposed surface at the remnant of a stem on Line Bl (not to scale).
We must now find a letter to which we can restore this stem. Since the
stem is situated close to the preceding zayin, and since it stands quite
upright, we can posit only four letters as candidates: gimel, heh, waw and
resh. When we compare the stem with the shape of gimel two letters
80 The Tel Dan Inscription
earlier, we see that the stem here is slightly longer than the right stroke of
the gimel. However, it is not so much longer as to make correspondence
impossible. The difficulty with restoring this stem to a gimel is philologi-
cal, as we would be hard pressed to interpret the word.
The heh is an unlikely candidate on epigraphic grounds. When we com-
pare the form of heh that occurs in Fragment B with the stem, we find that
the stem stands too upright to be considered a heh. Also, we would expect
to see a trace of at least the bottom 'bristle' of the heh on the extant written
surface. Thus, the stem is too anomalous to the form of a heh to be restored
as one.
A waw is more plausible, as the stem is on a comparable angle and in
the right position. The only factor of doubt is the length of the stem.
Although it is not outside the plausible scope of a waw, the stem of a waw
tends to be shorter rather than longer. The stem would require us to iden-
tify this as a long-stemmed waw.
The only letter that does not present such difficulties is resh. It is com-
parable with the forms of resh seen in Fragment B2. It also is philologically
easier to restore a resh here than a waw. As such, I recommend restoring
this stem as a resh.
The clear evidence gleaned from the vertical strokes of Line 1 suggests
that we have the engraver of this fragment in a comparable position with
that seen in Fragment Athat is, along the left edge of the original inscrip-
tion. The bottom of the restored resh is quite shallow, in keeping with an
engraving starting point. Thus, all the vertical strokes, with the exception of
the connecting diagonal in the zayin, show signs of having been carved
from bottom to top. The diagonal of the zayin is inconclusive. This can be
attributed to the shape of the letter in which the two horizontal strokes are
carved over the extremities of the diagonal connecting them. However, the
evidence of the other letters in Line 1 suggests this diagonal was also carved
upwards, probably as the first stroke of the letter.
Therefore, we may decipher Line 1 as: [...]iT^l[...].
Line 2
The first letter of Line 2 has been destroyed except for the remains of a
long stem that curves slightly upwards at the end. The stem is too close to
the next letter to restore it as a mem. A kaph or peh could possibly be
crammed in, while a nun or taw would fit with more ease. There is the
slightest evidence of a second stroke right on the edge of breakage. This
stroke is too irregular in shape and position to allow for a peh or nun. It
3. Epigraphical Analysis 81
does, however, comport well with either the left 'finger' ofakaph or the
leftmost part of a crossbar on a taw. Philologically, a taw is easier to
restore than a kaph. There is also slightly less room for a kaph than for a
taw, so I recommend restoring this stem as a taw.
Following this is the greater part of a lamed. Only the well-rounded
'hook' and the bottom half of the 'shank' have been preserved, making its
identification secure. After this, we find a heth. There is some slight chip-
ping along the middle 'rung', probably inflicted accidentally during carving.
Both vertical 'runners' have been carved from the bottom up. Following this
we have a mem and a heh which both display signs of having been carved
upwards.
A very small word divider follows next. There is some erosion on the
surface of the stone here, as well as a chip. After it, we see a very neat
angular beth. The spine does not curve into the tail as was the case with
most beths in Fragment A. Rather, the spine bends into the tale at a defi-
nite angle. Following this we have a neat 'aleph. Both diagonal strokes of
this letter have been carved upwards.
Next to this, we have the remains of a letter that has been almost totally
destroyed. The remaining part consists of a stroke running along the line
of breakage. A good candidate for this letter is beth as there may be a
slight bend upwards at the top of this stroke indicative of the bend between
the tail and the spine of the letter. Biran and Naveh observed that if this
letter is a beth, it sits abnormally high up on the line. They suggested apeh
with the note that it is abnormally straight for apeh.17 The roundness and
length of the letter peh means we should dismiss it as a possible candidate
for restoration.
The suggestion of beth, however, is not as troublesome as it first seems.
First, when considering this stroke as part of a beth, one is instinctively
drawn to comparing it to the beth two letters earlier. It must be said, how-
ever, that this preceding beth sits slightly lower than is usual for the beths
of Fragment B. The spine and tail of this beth also hang lower than the
other beths of Fragment B. As such, we must see this preceding beth as
extraordinary. When we examine the other beths of Fragment B, we see
that this stroke remnant in Line 2 is not sitting much higher than is to be
expected, though we must still regard it as slightly high. We may note a
parallel case in Line 12 of Fragment A where the two lameds sit at vastly
different heights on the line. In that case also, the first instance of the letter
was extraordinary and the second was more typical.
Having said this, however, we must observe that there are other candi-
dates for restoration of this letter. Biran and Naveh neglected to consider
yodh and sin. We note that the yodhs of Fragment B are of the angular
variety so that the tail of the letter is straight and mostly on a considerable
angle. They also tend to sit low as they do in Fragment A. A sin is equally
plausible. A less likely, but still possible, candidate is lamed. This option
requires us to view the 'barb' on the 'hook' of the lamed as fairly straight,
but it is within reason. Therefore, I suggest either beth,yodh, lamed QV sin
as possibilities for restoration, but we cannot choose between them on
purely epigraphic grounds.
The visible strokes on this line are carved in the same manner as those
of Line 1 and all the lines in Fragment A. That is, the vertical strokes were
all carved from the bottom upwards.
We can, therefore, decipher Line 2 as: [.. .?]K!>nEnl?n[...].
Line 3
The third line of Fragment B begins with the remains of a heh. Only the
very edges of the three 'bristles' are still extant. The only other possibility
for this letter is a samekh. However, we observe from Line 6 below that
the three crossbars of the samekh are a good deal closer to the horizontal
than the 'bristles' of any heh. For this reason, the crossbars of a samekh sit
fairly high on the line, whereas the 'bristles' of all hehs tend to droop low.
We observe the same tendency in Fragment A also. Therefore, we are
justified in rejecting this letter as a samekh and considering it as a heh.
Next to this we have what appears to be a word divider. It is so shallow,
however, that we may have doubts as to whether it is in fact a word
divider. It actually bears more resemblance to one of the small pockmarks
that fleck the stone of Fragment B. However, since it occurs in a position
where a word divider is expected, we may reasonably assume it to be one.
We then find the first fully preserved letter of this line, a waw. The
upward direction of carving is very noticeable on this letter. The 'hook' of
the waw is angular and fairly open. Next to this we have an angular yodh
that, due to erosion, is quite shallow. However, immediately above the tail
of the yodh is a small chip. This is presumably one of the characteristic
pockmarks on Fragment B. However, it comes close to passing as a word
divider. It occurs in the exact same latitude as the word divider preceding
the waw. However, since a word divider was identified before the waw, it is
highly unlikely that this is a word divider after the waw. Since both mark-
ings are similar, however, we must decide which is likely to have been the
3. Epigraphical Analysis 83
intended word divider. We could place the waw at the end of a word or at
the beginning of the next word with equal plausibility. A decision must
come down to the fact that there is more of a gap between the restored heh
and waw than between the waw and subsequent yodh. Thus, we should
maintain our view that there is a word divider before the waw and dismiss
the marking near the tail of the following yodh as a small pockmark or an
engraver's error. A similar pockmark is observable between this yodh and
the heth in the line above it.
Following the yodh is a neat 'ayin. Some erosion has blurred the con-
tours of the letter towards the upper right. Next to this we have a lamed,
followed by a small word divider. The next word begins with a mem that
displays a gently curving stem like the mem in Line 2. Following this is a
lamed and a very neat kaph. There is a slight gap to the last extant letter of
Line 3, ayodh. Some of this character has been broken off at the edge of
the fragment, but there is no doubt about its identification. From what is
visible of this yodh, it appears to be of the rounded variety rather than the
angular.
Again, the engraver appears to have incised the individual strokes in the
same manner as previous linesnamely, in an upward direction.
Line 3 can, then, be deciphered as: [...fD^E-^IH !![...].
Line 4
There has been much debate over what exactly lies at the beginning of Line
4 on Fragment B. When Biran and Naveh published Fragment B, they recon-
structed a yodh as the first letter, before the first fully legible letter, a heh.18
Indeed, their brief commentary took for granted that a yodh was easily
restored in this place for they drew no attention to their restoration what-
soever. From the photographs and facsimile of Fragment B, it became clear
that Biran and Naveh had based their restoration on two small markings on
the very edge of the fragment at Line 4. However, their confident recon-
struction provoked criticism from both Cryer and Thompson who main-
tained that there was no evidence of any yodh before the first heh on Line
4.79 They appealed to the photograph in Figure 4 of Biran and Naveh's pub-
lication as showing no traces of ajW/z.80 They therefore accused Biran and
Naveh of having 'added' a yodh to the beginning of Line 4.
78. Biran and Naveh, 'The Tel Dan Inscription', pp. 9, 12-15.
79. F.H. Cryer, 'King Hadad', 5/079(1995), pp. 223-35 (232); Thompson, 'Dis-
sonance and Disconnections', p. 238.
80. Biran and Naveh, 'The Tel Dan Inscription', p. 5, Fig. 4.
84 The Tel Dan Inscription
81. Cf. the difference in positions of the he and two daleths at the end of this line.
3. Epigraphical Analysis 85
the lower marking here in Line 4. Comparison with the preceding shows
the plausibility of suggesting a , also. The bottom marking would be the
remnant of the lowest 'bristle' of the with the other 'bristles' broken off
during fracture of the stone, but leaving the word divider partly preserved.
The same observations may be made for a samekh. Since there is nothing
requiring us to see the two markings here as parts of the same letter, it is
reasonable to suggest that one marking is a word divider and the other is
part of a letter. Indeed, if the upper marking is a word divider, it is cer-
tainly in a suitable position.
Another possibility is that the upper marking is a word divider and the
lower marking represents the end ofapeh.++++++++++++++++++
well as Fragment A that the peh begins on a similar angle to that on which
the lower marking here sits. One difficulty, however, is that for two of the
three clear instances ofpeh appearing on the Tel Dan fragments, the lower
extremity of the figure occurs further left than the starting point of the
figure.82 This being the case, we would expect to see some of the tail of the
peh still visible here on the fragment. No such remnant of the tail exists,
however. It is, therefore, only with great difficulty that we may suggest
restoring a peh here. We may also rule outazayin, for although the lower
marking is in an appropriate position for a zayin, the angle of the stroke
makes it highly unlikely.
As such, we see that Biran and Naveh's suggestion ofyodhis improb-
able. It is far more likely that the upper marking is a word divider83 and the
lower marking is the remnant of a gimel, or samekh.
As was mentioned previously, the first fully legible character on Line 4
is a . Interestingly, the incisions of the bottom two 'bristles' barely touch
the stem of the letter. The stem has also been carved a good deal deeper
than the other strokes. After this we have a mem with a stem so long that
it almost touches the qoph directly beneath it on Line 5. This is followed
by a lamed, kaph and a small word divider.
The next word begins with a that sits unusually low when compared
with the two daleths which follow. On the first daleth we observe a small
pockmark at the bottom right junction which makes the figure look as
though the horizontal stroke has extended past the right diagonal towards
the . This is not the case. The second daleth is somewhat squashed at the
leftmost point. At this point, the fragment breaks off. However, the con-
tour of the shape of the breakage here still leaves the remnant of one more
82. The only exception is the second peh on Line 7, Fragment A, in the word EHS.
83. Cf. Cryer, 'King Hadad', p. 232.
86 The Tel Dan Inscription
letter. The remnant consists of two strokes in the form of an 'X', which
can only be identified as the topmost portion of an 'aleph. Unfortunately,
all other markings on this line have been destroyed.
Consistent with the method of carving seen in other lines and in Frag-
ment A, the vertical strokes of this line have been carved upwards. This
confirms that the engraver was positioned on the left side of the original
inscription.
We may, therefore, decipher Line 4 as: [.. .]& "["!!"[ ~[u?Qn*[?...].
Line 5
Line 5 of Fragment B1 is the first to match up with a line in Fragment B2.
The first letter is clearly an 'aleph, followed by apeh, qoph and a very deep
word divider. After this divider, we have the most part of a mem. This let-
ter appears somewhat crowded and approximately half of the stem has
been destroyed by a lacuna in the stone. This lacuna imposes over almost
all of the following letter and extends downwards to the edge of Fragment
B1 along the join with Fragment B2.
The letter, which the lacuna almost completely covers, has only the
slightest remnants left. Along the left edge of the lacuna may be seen faint
traces of a vertical line between the same latitudes as the leftmost stroke of
the preceding mem. There is also some scarification next to this faint line,
giving the false impression that this line is connected to the word divider
which follows. This impression is not helped by the fact that the word
divider is not well carved and appears to have undergone slight chipping
to make it appear like a vertical line rather than a dot. Some of this letter's
stem is visible below the lacuna on Fragment Bl. These remnants are
enough to confirm the letter as a nun.
Following the elongated word divider is a sin. The line of breakage runs
diagonally across this letter from top left to bottom right. However, the
lower portion has been preserved on Fragment B2 and is therefore entirely
visible. Some abrasion and minor chipping has affected the letter slightly,
but not so much as to put its identification in doubt.
After the sin is a beth. This is the first letter contained wholly on Frag-
ment B2. We note that this beth is considerably smaller in size to the beth
encountered in Line 2 above. After this letter we have the right half of a
fairly eroded 'ayin. The left half has been destroyed by the breakage as
Fragment B2 ends here. We may discount the possibility of a qoph here
because we would expect still to see most of the stem on Fragment B2.
This, however, is lacking. Between the beth and the 'ayin is a small pock-
3. Epigraphical Analysis 87
mark that could easily be mistaken for a word divider. Only close exami-
nation reveals that it is an irregular shape for one. In addition, the fact that
there is no substantial gap between the beth and restored 'ayin counts
against it being a word divider.
No change to the engraver's position is seen in the nature of the strokes
of this line. Once again, the vertical incisions have been carved from the
bottom upwards.
Line 5, then, should be deciphered as: [.. .]iDC>jQpSK[...].
Line 6
The beginning of Line 6 starts on Fragment B1. The line of breakage here
has left only the smallest trace of a letter. It consists of one slightly curved
stroke that lies closer to the horizontal axis than the vertical. Biran and
Naveh took this for the edge of an 'ayin+++++++++++++++++++++
anything different. Indeed, the photograph of Fragment Bl would seem to
confirm this as a sensible suggestion.85
However, once again we find that the photograph is a little deceiving.
Oddly enough, the photograph of Fragment B1 exaggerates the axis of this
curved stroke so that it looks more vertical than it actually is on the stone.
The photograph of all three fragments arranged by Biran and Naveh exag-
gerates it even more so.86 The eye is also prompted to exaggerate the axis
by the broken edge of the fragment. Yet, this curved stroke is a good deal
more horizontal than the photographs suggest.
This gives the impetus to suggest more than just an 'ayin as a possible
restoration of this letter. Similar in shape would be the corresponding por-
tions of a teth, apeh or the head of either a qoph or ayodh. Epigraphically,
all these letters are just as plausible as an 'ayin. It is, however, on phi-
lological grounds that we must restore this letter as an 'ayin.
Following this restored 'ayin, we have the top half of a nun. Most of the
stem has been destroyed by breakage. Yet, the portion still visible reflects
a form of nun that is some way between the nun restored at the lacuna in
Line 5 and the nuns of Fragment A. The visible portion shows the con-
necting horizontal stroke of the nun curving significantly up to join with
the top of the stem. The only other reasonable alternative is to consider
this letter as a waw. However, this poses more epigraphical difficulties
than a nun.
84. Biran and Naveh, 'The Tel Dan Inscription', pp. 12, 16.
85. Biran and Naveh, 'The Tel Dan Inscription', p. 5, Fig. 4.
86. Biran and Naveh, 'The Tel Dan Inscription', p. 10, Fig. 9.
88 The Tel Dan Inscription
After this letter we see a word divider. The next letter is an 'aleph which
has its 'nose' broken off in the lacuna on Fragment Bl. The edge of the
lacuna follows the line of the letter's 'spine'. The next letter, a samekh,
occurs on Fragment B2. A small portion of the top two crossbars has been
eliminated by the breakage of the stone and the lacuna in Fragment B1.
Following this samekh we observe a resh and a yodh. After another
shallow word divider, the last extant letter on the line is an 'aleph, the
'nose' of which has been destroyed in the stone's breakage.
Despite the narrowness of the line, we may still discern the upward
direction of carving on individual strokes. Thus, the engraver was posi-
tioned at the left edge of the originally whole inscription during carving.
We may decipher Line 6, then, as: [.. .]K'HDfcj[...].
Line 7
The last two lines of Fragment B occur entirely on Fragment B2. The
beginning of both these lines is marked by significant abrasions and chips
over the surface of the stone. The first letter of Line 7 is a resh. It is
unfortunate that no characters before this resh on Line 7 have been
preserved, as the breakage of the stone here is quite shallow. The depth of
breakage before the resh is more typical of a lacuna than the edge of a
fragment. In depth, it compares with the lacuna of Line 5 on Fragment B1.
Thus, a thin segment of the stone's surface was chipped off in the
breakage of the stone, but it has not survived.
After this resh we see a mem, the head of which is quite cramped in
itself. The stem is very long and passes over an area of significant scari-
fication. A word divider follows, after which we see a beth and a resh. The
bottom tip of the beth runs into a pockmark and between it and the resh is
another characteristic small pockmark. A word divider follows the resh.
After the word divider we have what might be the possible remains of a
letter on the very edge of the fragment. The edge of the fragment is oddly
cut here, perhaps facilitated by the contours of this next letter's strokes.
This is certainly possible as the fragment itself is not very thick at this
point and is susceptible to chipping and breakage along any previous mark-
ing on the stone's surface. However, the nature of breakage here does not
allow us to distinguish whether there is in fact the smallest part of a stroke
here, or whether it is just part of the broken edge.
Schniedewind, however, maintains that there are two markings along
the edge of the fragment (see Fig. 3.19). Furthermore, he claims that these
two markings match exactly with the right extremity of an 'aleph, con-
3. Epigraphical Analysis 89
This phenomenon is not seen in any 'aleph attested on any of the three
fragments. Even among the straighter-sitting examples, the diagonal stroke
of every 'aleph invariably ends further left than the end of the horizontal
stroke. Thus, the only way Schniedewind can maintain the restoration of
an 'aleph is by massive clockwise rotation of the letter. Using electronic
imaging, it was found that a typical form of 'aleph as found in Fragment B
requires a clockwise rotation of a full 40 in order to match the contours of
the jagged edge at the end of Line 7.88 Even then, the letter sits abnormally
high on the line. Such quantitative reconfiguration cannot have any author-
ity. Schniedewind here would appear to be unduly influenced by con-
firming a match between the inscription's content and the biblical text.89
Schniedewind's proposal of two markings along the edge of the frag-
ment would better comport with a lamed. In such a case, there is no need
for repositioning or rotation of the letter. The fact that this more likely
possibility was not discussed by Schniedewind perhaps highlights the aim
of his reconstruction. Yet, it must be maintained that any restoration of a
letter after the word divider at the end of Line 7 is competely elusive. It
would be purely speculative to suggest any letter for consideration. We
must, therefore, decline any restoration here because the context is epi-
graphically too ambiguous.
It is abundantly clear that the vertical strokes of the letters on this line
have been carved from the bottom upwards. Deep points at the upper
extremities of these strokes confirm this, giving further confirmation to the
notion of the engraver being situated along the original inscription's left
edge.
Line 7, therefore, should be deciphered as: [.. .]~DD"l[...].
Line 8
The breakage along the bottom of Fragment B2 impinges upon the first
few extant letters of Line 8. The first letter is undoubtedly ayodh. Only the
'head' of the yodh remains extant and it is comparable to theyodh encoun-
tered in Line 6. The angle at which these strokes lie precludes the figure
from being a . After this yodh, the broken edge rises up slightly and
occupies the space between the yodh and a that follows. This has been
broken off below the bottom 'bristle'. There is, however, quite a
considerable gap between this and the preceding yodh. In fact, the gap is
so telling that we could reasonably expect a word divider to have been
here. However, no markings have been preserved between the two letters
because of the contours of the stone's breakage. As such, we must mark
the possibility of a word divider but cannot confirm one on epigraphic
grounds alone.
Following the is an odd figure that could be interpreted as either of four
letters: waw, kaph, nun or taw. Biran and Naveh understood this letter as a
waw without any comment as to its unusual form.90 It consists of a short
horizontal stroke joined to a vertical stem that kinks slightly at the
junction. Not all of this vertical stroke has survived. It would appear to
match a waw well except for the puzzling absence of a vertical stroke on
the 'hook'. We find such forms of waw in some Ammonite seals and bullae
from the seventh century.91 Yet, such a form is anomalous within the
epigraphic context of Fragment B because a distinctly Aramaic style waw
has already been observed in Line 3.
This leaves the figure open for interpretation. If it may be understood as
a waw with a missing stroke, we could similarly posit a kaph with a
missing 'finger'. The main objection to this would be the angle on which
the character sits. Likewise, if it were a taw with only half a crossbar, the
angle troubles us again. If we posit an incompleted nun, we must contend
with the fact that the stem does not appear to curve at all.
It appears that Biran and Naveh accepted this letter as waw to accommo-
date their reading of this line as part of a personal name with a theophoric
element, -yahu. However, a kaph would also make good grammatical sense
as part of the verb ~[iT, followed probably by the subject.92 In order to
derive a letter for restoration with any amount of certainty, though, we
should ask whether we can account for the fact that a stroke appears to be
missing without appealing to an Ammonite style of waw. Indeed, such an
appeal would appear out of place for fragments found at Tel Dan.
90. Biran and Naveh, 'The Tel Dan Inscription', pp. 9, 16-17.
91. E.g. the seals of 'Anamawt and Manahham. See W.E. Aufrecht, A Corpus of
Ammonite Inscriptions (Lewiston, NY: Edwin Mellen Press, 1989), 44, 48, respec-
tively.
92. This, of course, presumes that there is no word divider between the initial yodh
and at the beginning of this line.
92 The Tel Dan Inscription
The only letter for which we can discern such a reason, however, is in
fact waw. After this figure appears a word divider. However, between the
figure and this word divider is a small pockmark that occupies the approxi-
mate position where we expect the missing stroke of a waw to have been.
As such, it may have been the case that this pockmark in the stone existed
prior to the carving of the inscription. In that case, carving an entire 'hook'
may have led to further chipping or jeopardized the aesthetic integrity of
these figures. It may also have been the case that the engraver deemed this
pockmark adequate in itself for representing the vertical part of the 'hook'.
Indeed, when the inscription is viewed from a distance, this pockmark
does appear to act as such a stroke, albeit slightly detached. Thus, we have
good reason to presume this letter was intended as a waw. Although the
same arguments may be employed for seeing this figure as a nun, the prob-
lem of no curvature in the stem still remains, as well as the oddity of an
extended main stem. Therefore, we may recommend a waw for restoration
here.
After the word divider we encounter a slightly eroded beth that has
more in common with the beths of Fragment A than any of the other beths
in Fragment B. Despite the erosion, it is clearly visible that the 'tail' of
this beth is uniformly curved rather than bent. Following this is a resh. The
'nose' of the resh and the very bottom of the stem have been broken off.
Fragment B2 ends at this point.
Despite the paucity of full figures in the line, deep points are still obser-
vable on the majority of vertical strokes at the upper extremities. Only the
first letter, yodh, is ambiguous due to erosion. Thus, as was the case with
all the lines in Fragment A, the engraver of Fragment B appears to have
been situated along the left edge of the inscription.
We can decipher Line 8, therefore, as: [...]"D1!T[...].
Fragment B: Assessment
The identical carving characteristics of the inscribed characters on both
Fragment B and Fragment A suggest a strong connection between the two
fragmentary unitsnamely, that they were part of the same inscription. In
both Fragment A and Fragment B, clear signs that the engraver was
positioned along the original left edge of the inscription during carving can
be observed. However, some of the characteristics of Fragment B suggest
that it should not be placed to the immediate left of Fragment A. Rather,
since the letter neatness, the close lines of text, and the tendency to lift the
3. Epigraphical Analysis 93
slope of the lines all suggest that Fragment B was towards the end of an
inscription, there is a case to be made for Fragment B being placed below
Fragment A.
In accordance with the data calculated during the epigraphical analysis
of Fragment A, the dimensions of the original written surface of Fragment
B must be the same as that of Fragment A. That is, the original front face
of the inscription measured at least 110 cm in height and approximately 35
cm in width.
Having established that there is a very real epigraphical connection
between Fragment A and Fragment B, I turn now to a palaeographical
analysis of the script on each fragment. This analysis should elucidate any
connection even further.
Chapter 4
PALAEOGRAPHICAL ANALYSIS
Introductory Remarks
Having undertaken an epigraphical analysis of both Fragment A and Frag-
ment B, we are now in a position to examine the script. The epigraphical
analysis will have alerted us to the idiosyncrasies of the engraver. A palae-
ographical analysis will now allow us to explore the handwriting on the
inscription and place it within an approximate chronological bracket. From
this point, we will be able to offer a tentative date for the inscription to cor-
roborate with the archaeological context in which Fragment A was found.
Joseph Naveh, being an epigrapher and palaeographer, was criticized by
Cryer for not offering a full analysis of the script on the Tel Dan frag-
ments.1 Cryer was indeed justified in his remarks, for a script analysis is
fundamentally necessary before we can make any conclusions about any
inscription. Yet, Cryer's own examination of the letter types in Fragment
A is often very superficial,2 and his analysis of Fragment B is far from
comprehensive.
Cryer, however, was not the first to attempt a palaeographical analysis
of the script. In 1993, immediately after the publication of Fragment A,
Josef Tropper of Berlin offered an analysis of the script and included an
excellent table comparing the script of Fragment A with six other inscrip-
tions.3 However, his analysis of the script was confined to just nine letters.
Thus, a comprehensive analysis of the script of both fragments has not
been undertaken to date. It is my task to fill this notable void.
Amman Citadel inscription). For this reason, it has relevance for compari-
son with the Tel Dan fragments by virtue of the writing surface. However,
we must not be too hasty in giving it a great degree of authority because of
the nature of the handwriting. First, the text appears to have been etched
straight onto the limestone without a pre-written template. Also, the indi-
vidual letterforms within the text vary greatly from each other and are
actually quite messy. Thus, it is only with caution that we may compare the
script of the Tel Dan fragments with the letter forms in the Gezer Calendar.
Scripts gleaned from various seals and bullae, however, are of more
relevance. Although they are not public monumental inscriptions, they do
represent a type of formalized script. This being the case, they do parallel
the scripts of monumental lapidary inscriptions in writing conventions.
However, seals are physically very different from monumental lapidary
inscriptions. Thus, the scripts of seals and bullae are to be regarded with
greater significance than those of ink texts, but we must stop well short of
according them equal significance with other monumental lapidary inscrip-
tions.
The course of the present investigation will be to offer a thorough palae-
ographic analysis of each alphabetic letter attested in the fragments. Frag-
ment A will be examined independently of Fragment B in order to gain as
objective a picture of the script as is possible. Following this, I will exam-
ine the script of Fragment B,9 during which I will note any similarities or
differences with the script of Fragment A. By doing this, we will be able
to establish whether a plausible connection exists between Fragment A and
Fragment B in terms of palaeography.
Fragment A
A facsimile of Fragment A was drawn by Ada Yardeni soon after the
fragment's discovery (Fig. 4.1).
A-4 Line 4 ^1
The form is typically similar to a Latin 'A' fallen on its left side. The stem
extends beyond the two legs.10 The leftmost tip of the figure is generally
quite sharp and the triangular space enclosed by the three lines is quite
large and distinct. Common to each attestation is the slant of the stem, the
top leaning towards the left. There is also a tendency for the latter 'alephs
to bend the stem slightly leftwards above the highest point of intersection.
A further difference between the forms is the bottom horizontal stroke.
The angle of this stroke varies from the steepest of 17 to the horizontal
('aleph A-4), down to the flattest of only 2 ('aleph A-13)quite a large
disparity. The average slant is 11, corresponding to 'aleph A-8. The
stroke becomes generally flatter towards the end of the inscription. In
qualifying this, we must remember that each line of the fragment also
becomes flatter towards the end of the inscription. When this information
is combined with the angles of each incision of each 'aleph, a picture
emerges of the letters being gradually rotated clockwise throughout the in-
scription, in keeping with the general tendency of the lines to flatten. That
is, the rotation of the strokes in each 'aleph is very much influenced by the
slope of the line on which it appears. The first few 'alephs of Fragment A
appear rotated anti-clockwise because the lines slope considerably in that
direction.
The basic shape of the figure is observable in the ninth century BCE at
Tell Fakhariyah. However, not until the late ninth century BCE do we see
an attestation of 'aleph with a slant comparable to Fragment A. A prece-
dent for this form is seen in the Melqart Stele.'! There, we see a form with
10. 'Aleph A-6 has been slightly damaged near the left tip because of a vacuity in
the surface of the stone.
11. This is also known as the Bar Hadad Stele or the Bureij inscription. The precise
deciphering of the characters on the second line of this inscription has been a subject of
debate since its discovery at Bureij (near Aleppo) in 1939. A good summary of the
various interpretations can be found in W.T. Pitard, Ancient Damascus: A Historical
4. Palaeographical Analysis 99
Study of the Syrian City-State from the Earliest Times until its Fall to the Assyrians in
732 BCE (Winona Lake, IN: Eisenbrauns, 1987), pp. 138-44. Since Pitard's analysis
stems from an examination of the stele itself, I have taken his facsimile as the most
reliable, though it too has certain difficulties. See W.T. Pitard, 'The Identity of the Bir-
Hadad of the Melqart Stela', BASOR 272 (1988), pp. 3-22. For a photograph of the
inscription, see P.M. Cross, 'The Stele Dedicated to Melcarth by Ben-Hadad of Damas-
cus', BASOR 205 (1972), pp. 36-42.
12. See Donner and Rollig, Kanaandische und Aramdische Inschriften, I, 24.
13. P. von Luschan, Ausgrabungen in Sendschirli. V. Die Kleinfunde von Send-
schirli: Herausgabe undErgdnzung besorgt von Walter Andrae (5 vols.; Mittheilungen
aus den Orientalischen Sammlungen, Konigliche Museen zu Berlin, 15; Berlin: W. de
Gruyter, 1943), Taf. [Plates] 47.f, g; K. Galling, 'The Scepter of Wisdom: A Note on
the Gold Sheath ofZenjirli and Ecclesiastes 12:11', BASOR 119 (1950), pp. 15-18.
14. This archaic style is also seen in the later Azitawadda Inscription. However,
since we see the same type of 'aleph used in the Nora Stone, a Phoenician inscription
contemporary to the Kilamuwa Stele and Kilamuwa's Scepter, we can safely call this
an early style. Later Phoenician inscriptions do not use this type of 'aleph, further testi-
fying to its 'early' label.
15. W.F. Albright, 'New Light on the Early History of Phoenician Colonization',
BASOR 83 (1941), pp. 14-22 (15, Fig. 1).
16. Albright, 'New Light on the Early History', p. 18, Fig. 2.
17. Gibson, Textbook of Syrian Semitic Inscriptions, II, 5.
100 The Tel Dan Inscription
the figure from Zakkur is probably just idiosyncratic of the scribe rather
than the script as a whole.
Similar forms are also observed in Panammu I, the Sefire Treaties, the
inscribed bricks from Hamath18 and the Nerab Stele.19 By the end of this
period (the late eighth century BCE), different forms come into vogue. There
is a brief period of revival of the relevant form, though, in the late sixth and
early fifth century BCE in Phoenicia, when we observe comparable forms on
the sarcophagi of Tabnit20 and Eshmun'azar.
This information gives us a chronological horizon for the form of 'aleph
in Fragment A between the late ninth century BCE and the early fifth cen-
tury BCE. The form is attested in Sardinia, Cyprus, Southeast Anatolia,
Syria, Transjordan and later in Phoenicia.
Beth. There are 7 instances of the letter beth in Fragment A.21 Table 4.2
contains images of each of these in sequential order.
A-6 Line 7 23
A distinguishing trait of all these beths is their tail. The top of the stem
leans to the left and its tail curves uniformly underneath the head of the
figure. Beths A-1 and A-5, however, hint at the possibility of a form with a
vertexed tail. The tail is short, not protruding further left than the 'nose' of
the left-facing head. The head itself consists of straight strokes creating an
isosceles triangle that verges on the equilateral. As such, the head is of an
angular variety rather than a rounded one.
The beths of Northwest Semitic inscriptions may be put into two classes
according to the form of the letter's tail. The first form, under which we
may classify the beths of Fragment A, has a stem which curves neatly and
uniformly into a tail without any sign of a vertex between the stem and
tail. The second type, though, does display a vertex so that a definite point
can be assigned to where the stem ends and the tail begins. The tail of this
second type also displays little to no curvature. This second type is the
more common among monumental lapidary inscriptions. Examples of it
can be found at Tell Fakhariyah, in the Mesha Stele and Panammu I.
Although the curved-stem beth is attested earlier than the vertexed-stem
beth in Phoenician inscriptions of the tenth and early ninth century BCE,
there is a significant chronological overlap of both forms spanning centu-
ries. This is demonstrated well by Sefire Stele I A in which both styles
appear side by side. The fact that the vertexed-stem beth is the most com-
mon form beyond the tenth-century BCE Phoenicia, while both forms do
appear concurrently, suggests that the curved-stem beth is a variation of
the vertexed-stem form outside Phoenicia. In early Phoenician inscrip-
tions, however, the curved-stem beth prevails. This may suggest a definite
influence of Phoenician style over Fragment A since there is an almost
exclusive preference for curved-stem beths. In geographical terms, the
watershed of the Lebanon Mountain Range is all that separates Tel Dan
from Phoenicia, so it is certainly plausible. However, such a suggestion
would require Fragment A to be dated in the late tenth or early ninth cen-
tury BCE. From our examination of the archaeological context, we have
seen this cannot be the case.
Nevertheless, we must advise caution. It must be remembered that we
are dealing with only afragmen+++++++++++++++++++++++++
As such, what may appear to be a universal trend in a fragment might not
necessarily be a universal trend in a whole inscription. Thus, with regard
to the style of beth employed in an inscription, we should turn once again
to Sefire Stele I A. There, we see that not only are both styles of beth
employed side by side, but each style also occurs in a different place in the
102 The Tel Dan Inscription
22. The inclusion of the Amman Citadel Inscription as a 'Syrian' form affirms
Naveh's theory that Ammonite script was under the influence of Syrian conventions.
See J. Naveh, 'The Scripts in Palestine and Transjordan in the Iron Age', in J.H. Sand-
ers (ed.), Near Eastern Archaeology in the Twentieth Century: Essays in Honor of
4. Palaeographical Analysis 103
Bar Rakib.24 Both the sarcophagi of Tabnit and Eshmun'azar must also be
mentioned. To a lesser degree, we may mention the Sefire Treaties, the
Eqron Inscription25 and the stele from Tema.26 The figures in these three
inscriptions, although displaying similarly shaped forms, stand more up-
right than the beths of Fragment A.27
Thus, our chronological bracket for the beth of Fragment A ranges from
the late ninth century BCE right down to the mid-fifth century BCE. The
most meaningful matches derive from Southeast Anatolia, Syria and Trans-
jordan, though comparable forms are also found in Phoenicia and North-
west Arabia.
Daleth. Fragment A contains six instances of the letter daleth.2* Table 4.3
contains images of each in their sequential order.
The letter is the form of a triangle with the right diagonal extending a
little past the point where it meets the horizontal stroke. This horizontal
stroke is slightly curved and at first glance appears to vary considerably in
slant among the attestations, making it difficult to ascertain a typical form.
The differences are most obvious when comparing daleth A-3 with daleth
Nelson Glueck (Garden City, NY: Doubleday, 1970), pp. 277-83 (277-79); idem, Early
History of'the Alphabet, ^. 107.
23. This inscription is also known simply as the Panammu inscription. We have
chosen the label 'Panammu II' in order to distinguish it from the Panammu I inscrip-
tion. For treatment and photographs of the Panammu II inscription, see Gibson, Text-
book of Syrian Semitic Inscriptions, II, 14.
24. There are numerous fragmentary texts that we may ascribe to Bar Rakib. These
are referred to collectively here as 'Bar Rakib'. The label 'Bar Rakib Stele' will refer
to the first inscription of this small corpus, namely Bar Rakib A. See Donner and
Rollig, Kanaandische und Aramdische Inschriften, I, 216-21; Gibson, Textbook of
Syrian Semitic Inscriptions, II, 15-17,191, Fig. 11; Lidzbarski, Handbuch der Nord-
semitischen Epigraphik,+++++++++++++
25. S. Gitin, T. Dothan and J. Naveh, 'A Royal Dedicatory Inscription from Ekron',
IEJ41 (1997), pp. 1-16.
26. Gibson, Textbook of Syrian Semitic Inscriptions, II, 30.
27. Along the with the Sefire Treaties, the Eqron Inscription is a good example of
both the curved-stem and vertexed-stem beths appearing concurrently in one inscrip-
tion.
28. This is not including the daleth reconstructed in the damaged word at the begin-
ning of Line 2.
104 The Tel Dan Inscription
A-6. However, daleth A-2 has a slight chip at the horizontal stroke making
it look slanted down to the right in a way similar to daleth A-3, which is
immediately next to it in the inscription. Closer inspection, however,
reveals the point at which the horizontal stroke of daleth A-2 meets the
right-hand stroke to be higher than at first appears. It is, therefore, compa-
rable to daleth A-6. This means that daleth A-3 is anomalous and the rest
are quite similar to each other. Also of note is the leftmost tip, which
comes to a definite point. The characteristic feature is the rotation of the
entire letter such that it appears to lean towards the left.
Prior to the mid-ninth century BCE, a very upright form of daleth pre-
vails. In the mid-ninth century BCE, however, a rotation of the letter takes
place. This is observed in the contemporary inscriptions of the Mesha
Stele, which still preserves the upright form, the Melqart Stele and Amman
Citadel Inscription, which display the later rotated form, and the Archaic
Phoenician Inscription from Cyprus, which appears midway between the
two forms. Since Ammonite script shows definite influence from its more
northerly neighbours in Syria, this small survey shows that the rotated
daleth is a typically Syrian form. The idiosyncratic script of Tell Fakhari-
yah also displays a slanted daleth, but the shape of the letter differs in that
it has no tail at the bottom right vertex.
We notice another particular feature that is helpful in defining our time-
frame and area of influence for Fragment A. Whereas the form of Frag-
ment A displays a clear vertex at the bottom left of the figure, the Melqart
Stele shows definite curvature. The slightly later Amman Citadel Inscrip-
tion does display the vertex. The Horse's Ornament of Hazael, found on
4. Palaeographical Analysis 105
the Aegean island of Samos,29 displays the vertex but no rotation. Although
this particular inscription is not a monumental lapidary inscription, it is
useful in determining the nature of the form used in the third quarter of the
ninth century BCE. It appears that the rotation of the letter daleth occurred
during this time in the northern Levant so that by c. 825 BCE we see the
rotated form with vertex in the Kilamuwa Inscription. That this is defi-
nitely a Syrian form of daleth is furnished by evidence from the contem-
porary Nora Inscription, which still uses the Phoenician style of an upright
daleth, and the Syrian-influenced script of the Amman Citadel Inscription.
Thus, the Amman Citadel and Kilamuwa Inscription represent our terminus
a quo in c. 825 BCE.
The persistence of the curved daleth from the Melqart Stele is seen in
the eighth century BCE with the Zakkur Stele through to the Nerab Stelae
at the very end of the eighth century BCE. However, at the end of this cen-
tury BCE, we see a comparable form to that of Fragment A in the Siloam
Tunnel Inscription30 in Palestine. Among the Syrian forms, only the in-
scribed bricks from Hamath and the stele of Bar Rakib suffice for reason-
able matches in this century. After the eighth century BCE, other forms
dominate the register.
In addition to these observations, we may also note the conventions
used in non-dedicatory inscriptions. The Samaria Ostraca are particularly
useful here. Despite the convention to slant most of the letters to the right,
the ostraca preserve a form of daleth that is identical to that of Fragment
A. Ostracon 4 is a particularly good example.
Thus, this slanted daleth with three vertices and a tail has a fairly short
life between the late ninth century BCE and the end of the eighth century
BCE. The closest forms, however, occur between c. 825 BCE and c. 770 BCE.
This accords well with the approximate date gleaned from the archaeologi-
cal context of Fragment A. The form is attested in Southeast Anatolia,
Syria, Transjordan and Palestine.
Heh. The letter heh occurs six times in Fragment A. Table 4.4 contains
images of each in their sequential order.
29. I. Eph'al and J. Naveh, 'Hazael's Booty Inscriptions', IEJ39 (1989), pp. 192-
200.
30. S.A. Birnbaum, The Hebrew Scripts. II. The Plates (2 vols.; London: Palae-
ographia, 1954-57), 14; Lidzbarski, Handbuch der Nordsemitischen Epigraphik,II,
Taf. [Plates] XXI: 1.
106 The Tel Dan Inscription
A-5 Line 10 on
As with the other letters of the inscription, the heh sits on an anti-clockwise
slant. The figure is a type of'brush' or, as Cryer styles it, a 'comb'.31 The
three 'bristles' or 'teeth' project from the stem at an acute angle and tend
to be slightly shorter in length the further down the stem they are.32 The
stem is not particularly long either. Heh tends to be a long letter where the
stem extends quite low, but here in Fragment A, it is quite short. Some of
the stems also betray a minute curvature (see hehs A-1, A-3, A-4 and A-6).
The earliest match we have with this slanted form is the Melqart Stele of
Bar Hadad. The Nora Stone and the Archaic Phoenician Inscription from
Cyprus also display this form. In the Mesha Stele, the Azitawadda Inscrip-
tion and the Panammu II Stele, the heh is quite upright and the 'bristles'
are of equal length, although occasionally in Mesha, the middle 'bristle' is
shorter than the other two. As is to be expected, and in keeping with the
individual style of Mesha, the stem of the heh is also longer than that of
Fragment A. However, some hehs in the middle left of the Mesha Stele,
particularly between lines 15 and 26, show a clear affinity with the form in
Fragment A. Thus, it qualifies as a match. A slightly better proportioned
figure is seen in the Amman Citadel Inscription, though the stem is also
slightly longer than the stem in Fragment A.
Wow. There are nine occurrences of waw in Fragment A.33 Their images
appear in Table 4.5 in sequential order.
The waw is slightly slanted, as we have come to expect from Tel Dan.
The 'hook' part of the figure, which juts out from the top portion of the
stem, is comprised of two straight lines perpendicular to each other, simi-
lar to a Latin 'L'. However, in waw A-5 and waw A-7, the apex of the
'hook' is quite rounded. That portion of the stem that extends above the
'hook' is roughly of equal length to the parallel stroke of the 'hook'. The
stem is straight, although in some instances there is a decided kink at the
junction with the 'hook' (waw A-3 and waw A-8). This 'hook' style is typi-
cal of Aramaic or Syrian scripts as opposed to the open 'cup' style of more
southern scripts such as Phoenician and Moabite.
The distinctiveness of the waw allows us to make quite precise matches.
Our earliest attestations of the 'hook' form are the Archaic Phoenician
Inscription from Cyprus and the Kilamuwa Stele. The inscription from
Cyprus, though, shows traces of an earlier open 'cup' style, and so repre-
sents a terminus a quo. The waws of Kilamuwa tend to stand quite upright,
but the shape of the figure corresponds very closely to that of Fragment A.
The 'hook' of the figure is distinctively pointed and the variations of the
33. This is not including the reconstructed waw in the damaged word at the begin-
ning of Line 2.
108 The Tel Dan Inscription
figure throughout Kilamuwa's stele mirror the same variations that appear
in Fragment A. The same may be said for the stelae of Panammu I,
Panammu II and Bar Rakib.
34. This waw is clear on the stone but the image here is unfortunately unclear.
35. This waw is only partially preserved and not clear on the photograph. Thus, a
facsimile of the image has been employed instead.
36. W.L. Reed and F.V. Winnett, 'A Fragment of an Early Moabite Inscription
from Kerak', BASOR 172 (1963), pp. 1-9.
4. Palaeographical Analysis 109
Heth. There is only one instance of the letter heth in Fragment A.37 The
image appears in Table 4.6.
A-l Line 1 1 pN
37. This is not including the reconstructed heth in the damaged word at the begin-
ning of Line 2.
110 The Tel Dan Inscription
matrices. The one heth in line 7 and the last heth of line 17 are the com-
parable examples. The two heths in the Hazael Ivory are quite different
from each other. The first one affords some comparison in that the left
'riser' extends upward and the right 'riser' extends downward. The diffi-
culty with the heths of Kilamuwa is that they too vary in appearance. The
most common form has both tips of the left 'riser' extending considerably
beyond the 'rungs' while the right 'riser' extends beyond the 'rungs' only
at the bottom. Thus, it presents a similar, albeit disproportionate, shape to
that of Fragment A. The same may be said for the Siloam Tunnel.
The Zakkur Stele also has fewer constraints and yields the best matches
of all. The final heth of line 10 and the first of line 15 are perhaps the clos-
est figures. A good match with the 'risers' and slant of our heth is also
seen in the Amman Citadel Inscription. However, it represents a two-
'runged' form of heth, rather than the required three-'runged' form. This
two-'runged' form is characteristic of a southern style, so the variation is
attributable more to region than time. We do, however, find a two-'runged'
heth in Bar Rakib, but the slant and 'risers' of the Bar Rakib heth are not
comparable to the heth of Fragment A. Yet, since we see the two-'runged'
form as contemporary with the more common three-'runged' form, the dif-
ference cannot be chronological. Thus, the Amman Citadel must be in-
cluded as the best comparison to the heth of Fragment A.
All this data yields a date for our heth between the mid-ninth century
BCE and the early eighth century BCE, though we can plausibly posit a
spectrum down to the very end of the eighth century BCE. Topographically,
Syria and Transjordan afford the best comparisons, though a comparable
form is found in Palestine.
Yodh. The letteryodh occurs 16 times in Fragment A, more than any other
letter. Table 4.7 contains the images of all these yodhs in their sequential
order.
The yodh resembles a slanted Latin 'Z' or an Arabic numeral '2', with a
dash branching to the left from the vertical 'spine'. The dash is shorter in
length than the parallel top stroke. The vertices on the two ends of the
'spine' vary between angular (e.g. yodh A-3) and curved (e.g. yodh A-12).
The significant features of the yodh, however, are the angle and length of
the figure's 'tail' and the overall slant of the whole letter.
4. Palaeographical Analysis Ill
A-10 Line 6 -
A-15 Line 10 n:
Kaph. The letter kaph occurs eight times in Fragment A.41 Table 4.8
contains their images in sequential order of their appearance.
A telling feature of the kaph is its stem. It slopes down to the left, and
the bottom shows little or no curvature. A rotated ' V shape branches off
to the left from a point along the top half of the stem. This gives the figure
the appearance of three 'fingers' (the top part of the stem being one 'fin-
ger'). All three of these 'fingers' are of approximately equal length. The
angles between them are also approximately the same. This branching ' V
construction is most helpful in assigning a date and style to the form, as
there are numerous forms of kaph used in Northwest Semitic inscriptions.
It is, therefore, a distinguishing trait to aid our analysis.
The Phoenician Governor's Bowl from Cyprus has a very good match to
the kaph of Fragment A. Earlier, however, we see a comparable form in
the Nora Inscription, in which the ' V branches out from the mid-point of
the stem. Although the stem is considerably shorter and not at all curved,
the 'fingers' are of equal length, as in Fragment A. The kaph of Kilamuwa
39. Lidzbarski, Handbuch der Nordsemitischen Epigraphik, II, Taf. [Plates] II: 1.
40. The facsimile of Pitard in The Identity of the Bir-Hadad of the Melqart Stela'
reproduces the yodh of Line 4 as curved. However, a closer inspection reveals the yodh
to be quite straight and pointed, as reproduced by Cross in 'The Stele Dedicated to
Melcarth'.
41. Strangely, Cryer counts only six instances. See his 'Of Epistemology, Northwest-
Semitic Epigraphy and Irony: The "bytdwdftiouse of David" Inscription Revisited',
JSOT69 (1996), pp. 3-17 (14).
4. Palaeographical Analysis 113
is similar, except that the stem is notably curved. This is an occasional fea-
ture of the kaphs on the Mesha Stele. However, most of the forms in the
top half of the inscription are good matches. Kilamuwa's Sceptre is a bet-
ter match, being similar to the Nora Inscription. A votive inscription from
Kuntilet 'Ajrud42 also displays comparable characteristics.
A-2 Line 3 }
A-3 Line 5 }
A-5 Line? 32
A-6 Line 8 ^Q
A-7 Line 9 3
The slightly later figures from the Siloam Tunnel and the Eqron Inscrip-
tion have the left 'finger' slightly higher, as in kaphs A-l and A-2.
Thus, we have comparable examples for the Fragment A kaph dating to
between the mid-ninth century BCE and the early seventh century BCE from
Sardinia, Cyprus, Southeast Anatolia, Syria, Transjordan, Palestine and
even the Negev. This wide geographical spread is offset by the narrowness
of the assigned date, which spans a little over a century.
Lamed. The letter lamed occurs 15 times in Fragment A. Table 4.9 con-
tains the images of these instances in their sequential order.
A-8 Line 8 -t
44. The 'barb' and half the 'shank' of lamed A-3 have been damaged by a lacuna.
45. The top of lamed A-5 has been damaged.
46. Only the top part of lamed A-6 is visible on the inscription. The bottom portion
has been damaged. However, there is no other letter that could have a similar stroke.
The kaphs do not extend as high as this stroke suggests, the top stroke of the mem does
not have such an incline, and the taw never reaches as high. The lamed is easily the
tallest of all the letters and this visible stroke accords well with the attributes of the
Fragment A lamed.
4. Palaeographical Analysis 115
A-14 Line 12 ^
A-1548 Line 13 [...feu
The lamed is the tallest of all the letters appearing in the fragment. The
shape is a right facing 'fish hook'. The 'shank' of the 'hook' is on a signi-
ficant slant (between 32 and 43 right of the vertical) and the 'barb' curls
back up slightly. The actual curve of the 'hook' is not uniform. It bends
hyperbolically, sometimes achieving a quasi-vertex as in lamed A-8. The
'shank' is not completely straight, but rather displays a very subtle curve
and, sometimes, a slight kink as in lameds A-8 and A-ll.
The first matches may be made with some of the Phoenician inscriptions
of the tenth century BCE, such as the Ahiram Sarcophagus, and the inscrip-
tions of Yehimilk49 and Shipitba'al.50 In the ninth century BCE, the Mesha
Stele and El-Kerak Fragment contain lameds that have a comparably
slanted 'shank'. However, the curve of the 'hook' is quite regular and the
'barb' curls up towards the vertical. The Nora Inscription and the Archaic
Phoenician Inscription from Cyprus provide us with good parallels to the
lamed of Fragment A. They feature the hyperbolic curve that is charac-
teristic of Fragment A. The 'shanks' are also on a notable slant, and the
'barbs' curl up above the horizontal.
47. The top half of this letter has been damaged. The curved stroke is easily identi-
fiable as a lamed and fits the context best as a lamed.
48. A small portion of the lamed is missing due to the breakage of the stone.
49. Naveh, Early History of the Alphabet, p. 52, Fig. 43.
50. K. Jaros, HundertInschriften aus Kanaan und Israel (Fribourg: Verlag Schweiz
erisches Kamolishces Bibelwerk, 1982), 12.
116 The Tel Dan Inscription
51. Because this is only written in fired ceramic, it does not carry as much weight
for comparison with Tel Dan as do stone inscriptions. However, it does alert us to a
Phoenician style that undergirds the script of Fragment A. The decanter contains the
inscription binDrtib ('belonging to Natanba'al'), written on the shoulder of the decan-
ter, just below the neck. See R. Deutsch and M. Heltzer, New Epigraphic Evidence from
the Biblical Period (Tel Aviv: Archaeological Center Publications, 1995), pp. 39-40.
52. Jaros, Hundert Inschriften,30.
53. Jaros, Hundert Inschriften++++
54. B. Maisler, 'Two Hebrew Ostraca from Tell Qasile', JNES10 (1951), pp.
265-67.
55. A.Lemaire, 'Les Inscriptions de Khirbet el-Q6m',/?5 84 (1977), pp. 595-608.
56. The Ophel Ostracon. See S.A. Cook, 'Inscribed Hebrew Objects from Ophel',
PEFQS 56 (1924), pp. 183-86; Jaros, Hundert Inschriften,50.
57. H. Torczyner, Lachish. I. The Lachish Letters (4 vols.; The Wellcome Archaeo-
logical Research Expedition to the Near East, 1; London: Oxford University Press,
1938); O. Tufnell, M.A. Murray and D. Diringer (eds.), Lachish. III. The Iron Age (4
vols.; The Wellcome-Marston Archaeological Research Expedition to the Near East, 3;
London: Oxford University Press, 1953), pp. 21-23, 331-39.
4. PalaeographicalAnalysis+ +++
Mem. The letter mem occurs eight times in Fragment A.63 Table 4.10 con-
tains their images in sequential order of their appearance.
The form of the mem resembles a Latin ' W, weighed down on the right
by a stem. The strokes of this jagged head are of approximately equal
length and the stem curves under it slightly. The most significant feature of
the figure is the jagged head. The leftmost stroke always leans to the right.
On some figures, this stroke appears almost perfectly vertical. However,
this can be attributed to the slant of the line on which the particular letter
occurs. Thus, the jagged head is characteristically slanted to the right and
is quite large in proportion to the whole letter.
A-3 Line 6 ^B
A-7 Line 10 DH
64. The stem of this mem has been damaged by the major vacuity in the stone that
borders on the line of breakage.
65. This mem appears on the small extra fragment that fills part of the void in the
major vacuity. The stem has been completely damaged.
66. This mem begins the final extant line (13) of Fragment A. The lower portion of
the stem has been damaged in the stone breakage.
67. We may rightfully consider Shipitba'al the earliest example of a transition to
the horizontal head style since it is contemporary with other inscriptions still using the
vertical head style.
4. Palaeographical Analysis 119
Nun. There are only two instances of the letter nun in Fragment A, shown
in Table 4.11.
The two figures are a little different from each other in terms of the top
half of the letter. They both share a semi-curved stem on a significant
slant, but the cross and top strokes vary. In nun A-l, the cross stroke is
quite short and the top stroke leans to the left. Nun A-2 has a longer cross
stroke and a top stroke that leans to the right. The top stroke, in both
instances, is quite short relative to the stem and does not protrude below
68. A good example of this configuration is the Eqron Inscription. Bar Hadad's
Melqart Stele (mid-ninth century) is an early example of the horizontal configuration
of the jagged head. However, this configuration did not become the common style until
the seventh century.
120 The Tel Dan Inscription
the join with the cross stroke. Neither does the stem protrude above the
join with the cross stroke, which is almost horizontal. This gives it a 'zig-
zag' or 'lightning bolt' shape. Also, the join between the cross stroke and
the stem in nun A-2 appears to have a double vertex. Such a phenomenon
is quite unique among other attestations of nun in comparable inscriptions.
However, this double vertex is almost certainly a deviation from normal
style, rather than typical of it. Because it is such an isolated occurrence,
and nun A-l shows compliance with a major style, we can excuse the
double vertex as a 'mistake' by the scribe or engraver.
Slight changes occur in the writing of nun through the centuries. These
variations, subtle though they are, permit us to date the form on Fragment
A with a fair degree of accuracy. A precedent for this particular form is
seen in the inscription of Shipitba'al in the late tenth century BCE. How-
ever, this inscription still preserves a very rigid style rather than the neater
and slightly fancier style of Fragment A.
Bar Hadad's Melqart Stele, though, contains two nuns that are strikingly
similar to nun A-1. The Mesha Stele yields a comparable figure except for
the fact that the cross stroke is on a considerable slant. The Amman Cita-
del nun is a good match, as are the inscribed bricks from Hamath. The
Kilamuwa inscription displays nuns that compare with both nuns of Frag-
ment A. For example, in the word pi (Kilamuwa line 2), the nun is very
similar to nun A-l. Similarly, the word ]Sb> (Kilamuwa line 9) contains a
nun akin to nun A-2. It is interesting to note that the Kilamuwa Inscription,
while displaying Fragment A type variations, yields a typical form that is
somewhere in between the two nuns of Fragment A. This suggests that the
scribe (or engraver) of the Kilamuwa text had similar stylistic deviations
to the scribe (or engraver) of Fragment A.69
The nuns of Zakkur undergo a strange transformation. In Zakkur A, the
nuns are similar to the style of Fragment A, until we come to line 10, when
69. There is, however, no attestation of a double vertex in Kilamuwa, as in nun A-2.
4. Palaeographical Analysis 121
a different form begins to dominate. This form features a top stroke and
cross stroke that have merged into a significant curve, and it is this form
that Zakkur B displays. However, Zakkur still appears to be using the form
attested in Fragment A at Tel Dan. The Sefire Treaties also contain a wide
range of nuns, including forms similar to Fragment A.70 Panammu I,
Panammu II, and the Nerab Stelae also display similar forms.
Towards the end of the eighth century BCE, the styles of nun begin to
change slightly. The nun of Bar Rakib, for example, is similar to the form
in Fragment A, except that the cross stroke slopes downwards to the right.
This change is also evident in the Azitawadda Inscription and the Nerab
Stelae. Prior to this time, the form occurs only in the Phoenician influ-
enced enclaves of Cyprus and Sardinia, and so must be regarded as a
fringe form before the end of the eighth century BCE.
Cryer hesitantly suggested similarities with some of the Samaria
Ostraca.71 However, of all the Samaria Ostraca, only two or three examples
may be seen as comparable.72 All the other ostraca represent a style of nun
in which the stem extends above the join with the cross stroke, similar to an
open figure '4'a distinctly southern style of nun. Also, one of the com-
parable forms (from Ostracon 3) occurs with a more typical southern style
of nun, lending credence to the suggestion that the comparable nuns were
scribal 'errors'. Many of the inscriptions mentioned above also display
characteristic 'errors', but three things set them apart from the Samaria
Ostraca examples. First, Fragment A and the inscriptions mentioned above
do not display the southern style of nun at all. Their typical forms do not
project the stem above the cross stroke. A counter argument to this may be
that our sample of nuns in Fragment A is not large enough to be conclusive.
However, the second factor comes to bear on this point, namely, that the
Samaria Ostraca were written on sherds of pottery while Fragment A was
part of a stone stele. As was warned previously, pitting informal handwrit-
ing against formal handwriting can often lead to misjudgment. It is similar
to comparing one's own handwriting with typed or printed text. Different
implements are used, the writing surfaces differ, and the purpose of writing
and genre call for different styles of writing. Informal handwriting is more
likely to contain stylistic deviations than formal writing, which is carved
70. E.g. the two undamaged nuns in line 30 of Stele I, face B, lower portion, bear a
very close resemblance to Tel Dan nun A-l and nun A-2, respectively.
71. Cryer, 'On the Recently-Discovered "House of David" Inscription', p. 8.
72. These are from Ostraca 3, 4 and 5.
122 The Tel Dan Inscription
and given more thought, care and time.73 The mean deviation from typical
styles in the stone inscriptions, as well as the ostraca, suggests different
typical forms. The implication is that the three examples from the Samaria
Ostraca are far from concrete matches to cite with regards to the nuns of the
Tel Dan fragments. Despite the fact that the Samaria Ostraca are contempo-
rary with some of the above inscriptions, they should not really be men-
tioned in reference to the letter nun.
Thus, when all aspects are analysed, our chronological horizon for the
nun of Fragment A ranges from the mid-ninth century BCE to the late
eighth century BCE. Topographically, the style is concentrated in Syria, but
also present in Southeast Anatolia and Transjordan.
73. Needless to say, each case must be judged on its own merits.
4. Palaeographical Analysis 123
typically early style that separates the head of the figure into three individ-
ual crossbars. This style of figure is datable to between the tenth and
seventh centuries BCE. After this time, we see 'cursive' samekhs in use, in
which the three crossbars have evolved into one continuous 'zigzag'
construction. Yet, we may even refine this date for our samekh further by
virtue of the fact that the samekh of Fragment A has quite a long stem.
With these parameters, we can now look for comparable matches to our
samekh. The Mesha Stele provides excellent examples comparable to
Fragment A in all proportions. The stem, however, of Mesha is not slanted
at all and sits rather straight. The same may be said of the samekhs in the
Kilamuwa Stele. However, the samekh engraved on Kilamuwa's Sceptre
shows a definite slant. This slanted style of samekh has a short lifespan
mostly attested in scripts of a Syrian provenance or influence. In addition
to Kilamuwa's Sceptre, other accurate matches, such as the Amman
Citadel Inscription, Panammu I and the Azitawadda Inscription, demon-
strate this point. The samekhs of these inscriptions are proportionate to
Fragment A in their strokes and comparable in their slant. They are in all
respects excellent matches, which fall within a chronological bracket of
approximately one century. In the same time frame, the inscribed bricks
from Hamath may be mentioned. Though the crossbar there are much
shorter than those in Fragment A, the letter is on a comparable slant and so
constitutes a good match. Similar comments are reserved for Zakkur B.
Although we see the same slant in the form of samekh on the Silwan Tomb
Inscription,74 that samekh is disproportionately larger than the form in
Fragment A. It therefore cannot count as an accurate match.
By the end of the eighth century BCE, the form alters slightly, so that the
stem is only as tall as the bottom crossbar. With these data, we see that the
closest matching forms to the samekh of Fragment A are datable between
the second half of the ninth century BCE and the third quarter of the eighth
century BCE. However, we cannot discount a wider bracket between the
late tenth century BCE and the end of the eighth century BCE. The style is
attested over the entire Levant, as well as Cyprus.
'Ayin. The letter 'ayin occurs three times in Fragment A. Table 4.13
contains their images in their order of appearance.
74. N. Avigad, 'The Epitaph of a Royal Steward from Siloam Village', IEJ 3
(1953), pp. 137-52.
124 The Tel Dan Inscription
A-l Line 1 [-
A-2 Line 12 ++
The form of the letter is a simple closed circle. What appears to be a char-
acteristic 'bump' in the bottom left segment is the starting point at which
the engraver began carving the letter. Thus, this 'bump' is not necessarily
indicative of the actual style employed. For all intents and purposes, the
figure is a round circle. Cryer observed the figure as 'a right-leaning slight
oval'.75 However, this lean is observable only in 'ayin A-2. The other two
'ayins display no such rightward lean. Thus, an attempt to date the form
used here in Fragment A based on the figure's lean could be misleading.
It is certainly easy to read too much data into such a simple figure. Due
to the very plain nature of the letter, a date and topography is very difficult
to assign. We can go back as far as the Proto-Canaanite script of the
thirteenth-twelfth centuries BCE to find a match, down as late as the fifth
century BCE Eshmun'azar Sarcophagus from Sidon, and as far afield as
southern Arabia with Proto-Arabic script. This would give us a palaeo-
graphical spectrum of eight centuries and a geographical spread across the
entire ancient Near East.
Yet, we must limit our search to Northwest Semitic inscriptions. Two
notable characteristics may be mentioned for the Fragment A 'ayin which
will aid us in narrowing our chronological spectrum for the letter. The first
is the lack of a pictographic dot at the centre of the figure.76 In the early
eleventh century BCE, we still find this central dot in use on a fragment
77 78
from Nora. However, inscribed arrow-heads found in the Lebanon,
dating to later in the same century, show this pictographic style of dotted
'ayin dropping out of usage to be replaced by the simple circle 'ayin,
which we see used in Fragment A. By the tenth century BCE, the picto-
graphic 'ayin had virtually disappeared.79 Thus, a terminus a quo in the
tenth century BCE is appropriate. A terminus ad quern should be sought in
the fifth century BCE, in accordance with the script on the Eshmun'azar
Sarcophagus.
Our second notable characteristic allows us to slightly refine this wide
chronological bracket, and that is the overall size of the letter. The circle of
the 'ayin is actually quite small. It is no bigger than the heads of most other
letters. This feature starts to appear in the late tenth century BCE with the
Gezer Calendar. Byblian inscriptions still tend to produce large 'ayins at
this time, and even the Melqart stele displays large forms. It is only in the
mid-ninth century BCE, with the Mesha Stele, and the slightly later Amman
Citadel Inscription and Archaic Phoenician Inscription from Cyprus, that
we see the exclusive use of smaller 'ayins. Other inscriptions also employ-
ing the smaller form are Panammu I, Panammu II, the Nerab Stelae, the
Siloam Tunnel, the Silwan Tomb, the Eqron Inscription and the sarcophagi
of Tabnit and Eshmun'azar. Thus, although our terminus ad quern remains
in the early fifth century BCE, our terminus a quo has come down to the mid-
ninth century BCE, with the form attested over the entire Levant, as well as
Cyprus.
Peh. The letterpeh occurs twice in Fragment A. These instances are shown
below in Table 4.14.
The figure has no sharp anglesand that is its distinguishing feature. It
is a single curving stroke that faces concavely to the left. The top left of the
figure begins in an upward direction, but it then curves hyperbolically
downwards. It then straightens somewhat before beginning to curve sharply
Scripts', in P.M. Cross (ed.), Symposia Celebrating the Seventy-Fifth Anniversary of the
Founding of the American Schools of Oriental Research (2 vols.; Cambridge, MA:
American Schools of Oriental Research, 1979), I, pp. 97-123 (103-104); Naveh, Early
History of the Alphabet, pp. 40-41.
78. P.M. Cross, 'The Origin and Early Evolution of the Alphabet', 78 (1967), pp.
8-24 (21, Fig. 4); idem, 'Newly Found Inscriptions in Old Canaanite and Early Phoeni-
cian Script', BASOR 238 (1980), pp. 1-20 (4-6, Figs. 3, 5, 8); Naveh, Early History of
the Alphabet, pp. 38-40, Figs. 32-34.
79. There is the odd exception to this, the most notable being the Tell Fakhariyah
inscription. However, the script of Tell Fakhariyah is quite idiosyncratic.
126 The Tel Dan Inscription
again. It is at this point that the stroke finishes. It resembles the figure of a
lamed turned upside-down. The characteristic feature of this peh is the
concavity of the entire figure, and the fact that there are no vertices. Peh
A-l does almost reach a vertex at the top right, but it is still distinguish-
ably curved. Peh A-2 shows no signs whatsoever of a vertex, alerting us to
the certainty of an intentional curve.
As Table 4.15 shows, the letter is quite damaged due to a vacuity in the
face of the stone. Nevertheless, certain strokes of the letter can be clearly
distinguished and, with the clues of context, form the letter sadhe without
doubt. Clearly definable are the top strokes of the letter, which form a 'zig-
zag' step that descends slightly towards the right. On the far left is a
vertical stem that the vacuity has only partially damaged. This stem ex-
tends above the join with the 'zigzag' strokes, but how far it extends down-
wards is impossible to say, for both the vacuity and the breakage of the
stone have severed it. Further detail on the 'zigzag' strokes cannot be
gleaned because of the damage. If another sadhe was extant in Fragment
A, then at least some comparison could be made. It is, therefore, unwise to
speculate about how the figure probably looked when we have no pre-
cedent to go by.
This also means that comparison with other inscriptional sadhes is
meaningless. Despite the certainty that the letter is a sadhe, too many
doubts about its form exist to be able to compare it with other forms. Any
comparison would only be speculative, so judgment must be reserved.
Qoph. There are seven instances of the letter qoph in Fragment A. Their
images are contained in Table 4.16.
The form consists of an oval-type ring that is cut in two by a long stem.
The oval ring is on a slant with its longer axis running from bottom left to
top right. The top of the oval shows a slight but distinct 'dip' in the regular
curve in most of the forms. This 'dip' occurs at the point where the stem
meets the oval and is most distinct in qoph A-7. Qoph A-3 shows no sign
128 The Tel Dan Inscription
of a 'dip', while qoph A-l and A-6 are too damaged to tell. The remaining
four qophs, however, each have this dip. The stem does not extend above
the oval and the midpoint of the stem is at approximately the bottom
puncture of the oval. The stem is also generally straight, qoph A-5 being
the only example to display curvature.
This basic shape of the figure is in keeping with an early stage in the
development of the letter qoph. The same shape is seen in the Yehimilk
Inscription. However, most significant for dating the form in Fragment A
is the overall slant of the letter. This leftward leaning figure is first seen in
the mid-ninth century BCE on the Melqart Stele of Bar Hadad. It is also
seen in the Archaic Phoenician Inscription from Cyprus, Hazael's Horse
Ornament from Samos, the Zakkur Stele and the Sefire Treaties. The qoph
of Mesha is quite different in that it stands upright and is 'bow and arrow'
shaped,80 rather than an oval with a stem. It is thus an inappropriate match,
despite Cryer's comments to the contrary.81 The inscriptions of Kilamuwa,
Panammu I, Panammu II, Azitawadda and the grave stele of Si'-gabbari
from Nerab, all display qophs of a subtle but discernable leftward slant
and so constitute matches. A very good comparison is gained from Kila-
muwa's Sceptre. In that small gold engraving, we even see the 'dip' at the
80. That is, the 'arrow' is represented by the stem and the 'bow' is represented by
the figure's head or top.
81. Cryer, 'On the Recently-Discovered "House of David" Inscription', p. 8.
4. Palaeographical Analysis 129
top of the figure. The Eqron Inscription is the last attestation of this form
with such a slant. This is surprising for Palestine had developed a different
form ofqoph that is otherwise used universally.
Thus, despite the attestation in the Eqron Inscription, the formofqoph
used in Fragment A is of a distinctly northern style. We may place it within
a time-frame ranging from the mid-ninth century BCE to the early seventh
century BCE, with matches found in Cyprus, Southeast Anatolia, Syria and
Palestine.
Resh. There are eight instances of the letter resh in Fragment A. Their
images are contained in sequential order in Table 4.17 below.
A-2 Line 4 ^1
The figure consists of a long, left-leaning stem with a triangular head. The
top portion of the stem forms one of the skeloi of the triangular head and
bends slightly to the left at the bottom right vertex of the head. The mid-
point of the stem is at the approximate position of the head's bottom right
vertex, or slightly lower. The angles within the triangular head vary con-
siderably from one example to the next. With this fluctuation, it is difficult
to tell whether the bottom stroke of the head was intended to be com-
pletely horizontal or sloping slightly down to the right. The far left vertex
of the head is somewhat rounded in some of the examples, such as in reshs
A-2 and A-7, as is the top vertex in reshs A-4 and A-6. In addition, the
130 The Tel Dan Inscription
area of the triangular head varies noticeably from the largest (resh A-3) to
the smallest (resh A-6).
Precedents for the resh of Fragment A can be seen in the tenth century
BCE with the inscriptions of Yehimilk and Eliba'al82 from Byblos, and the
Gezer Calendar. At this stage of development, however, the leftward slant
of the letter had not yet been achieved. It is not until the mid-ninth century
BCE with the Melqart Stele that we find our first appropriate match. The
Melqart Stele displays two very different forms of resh. The first has a
pointed head, as in the Amman Citadel Inscription, and the second has
a rounded head, as in Zakkur. The first pointed form is quite adequate as a
match. The reshs in both the Mesha Stele and Amman Citadel Inscription
possess slightly longer stems than those of Fragment A, and the heads
have no curvature whatsoever. The one undamaged resh in the Amman
Citadel Inscription has comparable rotation, and the other damaged resh,
from what is visible, appears to follow suit. The forms in Mesha stand
quite upright, though there are many figures with a comparable slant. Kila-
muwa's Stele and Sceptre, as well as the Archaic Phoenician Inscription
from Cyprus, also deserve mention. The Nora Inscription has comparable
forms, but the slant of the figures speaks against a match.
The Sefire Treaties contain a variety of reshs, many of which correspond
with the forms in Fragment A. The vast majority of the Sefire reshs,
though, are quite pointed, without curved vertices. The examples that do
curve do so in such a manner as to obliterate any vertex at all, usually the
leftmost vertex. Nevertheless, the Sefire examples are comparable. The
resh of Panammu I is quite narrow and the strokes of the head are prone to
curvature. However, there is a reasonable correspondence, as with Panam-
mu II and Bar Rakib. Some of the forms from the Siloam Tunnel Inscrip-
tion seem comparable, but they are anomalous within the inscription itself.
The Siloam resh possesses a very long stem and a very curved leftmost
vertex on the head, rendering it inappropriate as a match. This is the case
with all forms attested in Palestine. Only the Eqron Inscription shows a
form midway between the typically long figure of Palestine and the shorter
form attested elsewhere.
The terminus ad quern should be put in the mid-fifth century BCE with
the Tema Inscription. Just prior to this time, we see comparable forms of
resh on the sarcophagi of Tabnit and Eshmun'azar. Thus, our spectrum for
the date of the resh goes from the mid-ninth century BCE to the mid-fifth
century BCE. The form is attested in Cyprus, Southeast Anatolia, Syria,
Transjordan and Northwest Arabia.
Sin. There are six instances of the letter sin in Fragment A. Table 4.18
contains their images in sequential order of their appearance.
The form of the sin is like that of a Latin ' W. The figure is wider than it is
tall and is on the slightest of clockwise rotations (6).83 All three vertices
are sharp and all four strokes are straight. None of the strokes overrun any
of the vertices. The figure also has two subtle nuances that will aid us in
dating it and comparing it to the handwriting of Fragments Bl and B2.
First, the right 'V of the figure is slightly larger than the left 'V. Second,
the left 'V of the figure tends to be rotated slightly more clockwise than
the right 'V. This is demonstrated by a comparison of the angles on which
the figure's 'parallel' strokes lean. The two right-leaning strokes are more
disparate than the left-leaning strokes. The right-leaning stroke on the right
' V stands more upright than the corresponding right-leaning stroke of the
left 'V'.84 Similarly, the left-leaning stroke on the left 'V stands more
upright than the corresponding left-leaning stroke on the right 'V, though
83. This is the average slant ofreshs A-2, A-3 and A-4. Reshs A-l, A-5 and A
are unable to be measured due to damage in the stone.
84. This phenomenon cannot be attested on sins A-1 and A-5 because of damage to
the stone.
132 The Tel Dan Inscription
85. This phenomenon cannot be attested on sin A-5 because of the large lacuna. It
is also impossible to verify on sin A-6 because the breakage of the stone has left no
trace of the left-leaning stroke of the left 'V.
86. This was in distinction to the leftmost stroke of the mem, which leaned to the
right.
87. A very slight erosion of the stone has blurred the sharpness of some strokes in
the sin of the Amman Citadel inscription. Nevertheless, the superficiality of the erosion
and the depth of the stroke cuts has ensured our ability to confirm the figure's prop-
erties.
4. Palaeographical Analysis 133
the Nerab Stelae all contain a number of sins of varying qualities. Some of
them do approach the form of Fragment A, allowing for respective matches.
The Kilamuwa Stele must be struck from a list of matches because it is a
tall and narrow figure that often displays curving strokes, especially in the
rightmost stroke. The Siloam Tunnel Inscription comes very close to Frag-
ment A in form, but not in rotation, whereas the Silwan Tomb Inscription is
close in rotation but slightly different in proportional form.
Thus, we observe a chronological horizon between the mid-ninth cen-
tury BCE and the end of the eighth century BCE for the Fragment A sin. The
matches are spread across Sardinia, Cyprus, Syria, Southeast Anatolia,
Palestine, Transjordan and even Mesopotamia.
Taw. There are five instances of the letter taw in Fragment A. Table 4.19
contains their images in sequential order.
The form is simply two strokes in the shape of a cross, rotated clockwise
between 58 (taw A-2) and 65 (taw A-3).89 The cross stroke is approxi-
mately perpendicular to the stem, intersecting it above the stem's midpoint
(except for taw A-2 in which it intersects at the stem's midpoint). In the
final two instances, the stem displays a slight kink at the point of inter-
section.90 Furthermore, taws A-l, A-3 and A-4 show minor signs of curv-
ing at the tail of the stem. In all instances, the cross stroke shows no
88. Only the bottom portion of the stem of this taw has been preserved.
89. Taw A-l does not preserve the cross stroke.
90. Breakage of the stone at the point of intersection precludes observation of this
trait in taw A-3.
134 The Tel Dan Inscription
curvature or bend and measures a little under half the length of the stem on
average.91
Two styles of taw are attested before the fifth century BCE: an equilateral
form and a long-stemmed form. The taws of Fragment A correspond with
the latter style since the stem is considerably longer than the cross stroke.
The difference in the two styles, however, appears to be regional. The
equilateral form is found in Phoenicia, Palestine and Transjordan while the
long-stemmed form is attested in Syria, Southeast Anatolia and Cyprus.
We note that the Amman Citadel Inscription displays forms of taw which
have affinity with both styles because the region is on the border of the
two styles. Tell Fakhariyah, which displays equilateral taws, is an excep-
tion to this geographical spread. However, since Tell Fakhariyah has idio-
syncratic forms of many letters, and is on the Northwest Semitic fringe,
the exception is accounted for.
Thus, we see a decidedly northern or Syrian influence over the taw of
Fragment A. Among these northern inscriptions we find some very close
matches. One of these is Bar Hadad's Melqart Stele. The figure there is
more upright than Fragment A, and the cross stroke does measure more
than half the stem. However, the essential shape with some rotation is
clear. Also of note are Hazael's Ivory Inscription, his Horse Ornament
from Samos, and the Archaic Phoenician Inscription from Cyprus. The
taws of Kilamuwa are quite fanciful, displaying a scimitar shape in which
the curvature of the stem is most obvious. The stem is also a good deal
longer than it is in Fragment A. It is less of a match than the previously
mentioned inscriptions, but it passes as a variation on a similar form.
The inscribed bricks from Hamath also deserve mention. The cross
stroke in these inscriptions is characteristically short on the left side. We
observe a similar phenomenon in taw A-5. A similar, but more rotated
form is observable in the Azitawadda Inscription. The taws in Zakkur's
Stele are quite close to those of Fragment A, though the cross stroke inter-
sects the stem quite high. Panammu I and the Nerab Stelae show the same
tendency. Panammu II, however, has the intersection much lower, as in
91. This is a calculation based on the average of the ratios between the stem and the
cross stroke in each taw except for taws A-l and A-3. Due to the position of the
breakage, full measurement of these two figures is not permitted. However, the lower
portion of the stem and the right portion of the cross stroke are still visible on taw A-3
and are able to be measured. The ratio of these two portions matches exactly the ratio
of the corresponding portions in taw A-5. Thus, despite not tallying taw A-3 in the
average calculation, the visible portions confirm the calculation of the other figures.
4. Palaeographical Analysis 135
taw A-2 of Fragment A. The Sefire Treaties are a good example of vari-
ations on the letter taw. A glance at Stele I B (a) epitomizes this with a
range of forms clearly observable. However, it is demonstrably clear that
the usual form of taw in the Sefire Treaties is the long-stemmed variety,
most of which are very close to the taws of Fragment A. Bar Rakib also
yields forms that are remarkably similar to those of Fragment A.
Considering these factors, the taw of Fragment A ranges chronologically
from the early ninth century BCE down to the end of the eighth century BC
and perhaps slightly later.
Fragment B
A facsimile of Fragment B (B 1 and B2 joined together) was drawn by Ada
Yardeni (Fig. 4.2).
B-5 Line 6 [-
138 The Tel Dan Inscription
The form of the figure is very similar, if not identical, to the form of 'aleph
observed in Fragment A. The forms are particularly close to the 'alephs oc-
curring in the latter portions of Fragment A. In particular, 'alephs A-7 and
A-10 show a near-exact correspondence with the regular forms in Fragment
B. This suggests a very strong link between Fragment A and Fragment B.
Of particular interest is the tendency to kink the top of the stem some-
what leftward. This idiosyncrasy was noticed in Fragment A also. Thus, it
is quite reasonable to suggest that the same hand lay behind both frag-
ments. Yet, caution is needed for we see the same peculiar kink in other
inscriptions, most notably the Amman Citadel Inscription and the Sefire
Inscriptions. Thus, the slight kink in the stem may not be characteristic of
an individual's writing technique, but of a prevailing style. Since it is such
a small detail, however, it is difficult to determine which is the case.
Since there is such a close match between the forms of 'aleph in both
fragments, all the considerations of the form in Fragment A apply here
also. Thus, we find matches in the Amman Citadel Inscription, the Archaic
Phoenician Inscription from Cyprus, the Nora Stone, Zakkur, the Sefire
Treaties, the Hamath Bricks, the Nerab Stelae and the sarcophagi of Tabnit
and Eshmun'azar.
Thus, the 'aleph of Fragment B certainly sits within the same time-frame
as the form in Fragment A. We may assign a date to the form between the
late ninth century BCE and the early fifth century BCE. The style is attested
in Sardinia, Cyprus, Syria, Phoenicia and Transjordan.
Beth. There are four instances of the letter beth in Fragment B. Their
images are shown in Table 4.21 in sequential order.
possibly the Tema Stele. Between these dates, we see particular affinity
with Kilamuwa, Nora, the Amman Citadel Inscription, the Archaic Phoe-
nician Inscription from Cyprus, Panammu I and the Sefire Treaties. The
Zakkur Stele has comparable forms except for a rounded 'nose'. Panammu
II, Bar Rakib and the Nerab Stelae represent a shift in style to a more up-
right letter, but nonetheless yield some good examples. Thus, the latter
inscriptions of Eshmun'azar and Tema are seen to bear archaic features
rather than contemporary features. For our chronology, however, we must
include them. Other figures, such as the more southern inscriptions like
Mesha and Siloam, display the same basic shape but lack the desired slant
of the whole letter. This shows the style spread across Sardinia, Cyprus,
Southeast Anatolia, Syria and Transjordan.
both display the same gimel with correct proportions and rotation. After
this time, good matches can be made with Kilamuwa, the Archaic Phoeni-
cian Inscription from Cyprus, Nora, Panammu I, Panammu II and Bar
Rakib. Unfortunately, gimel is unattested in the Melqart Stele and the Am-
man Citadel Inscription. The Mesha Stele and later inscriptions from Pales-
tine are all of a southern style, which is much narrower and longer than the
style of gimel here in Fragment B. Thus, Fragment B here shows a nor-
thern influence. The Sefire Treaties preserve a number of various gimels,
some of which comply with the form in Fragment B. The sole example
from Nerab tends towards the style of Fragment B also. On the other hand,
Azitawadda's Inscription is very similar to the southern style of gimel. The
sarcophagus of Tabnit from the late sixth century BCE should also be
mentioned. The Tema Stele from the mid-fifth century also shows some
affinity, though it is not as close as some of the previously mentioned
inscriptions.
Thus, we may date the form of gimel to between the late tenth century
BCE and the mid-fifth century BCE. However, most of our matches come
from the late ninth century and eighth centuries BCE. This is in line with
the evidence of the archaeological context of Fragments Bl and B2. The
form is attested in Sardinia, Cyprus, Southeast Anatolia, Syria, Phoenicia
and, to a lesser extent, in Northwest Arabia. This wide chronological and
geographical horizon can be attributed to the very simple nature of the
letter gimel.
Daleth. There are two instances of the letter daleth in Fragment B. Both
occur next to each other in the fourth line of Fragment Bl. Table 4.23
contains both their images.
The figure is essentially that of a triangle with a small tail. The tail is a
short continuation of the right diagonal stroke down below the bottom
horizontal stroke. Daleth B-1 also appears to have another minute tail with
the bottom horizontal stroke extending fractionally past the right stroke,
much like the corresponding portion of an 'aleph. However, such is not
142 The Tel Dan Inscription
the case. This 'tail'-like marking is simply a tiny pock in the surface of the
stone. The left and bottom strokes are of approximately the same size and
are both slightly longer than the right stroke.
There is a slight difference between the daleths of Fragment B and those
we observed in Fragment A. The daleths of Fragment B are proportionally
a little wider. This is shown by the bottom stroke being longer, and the
angle of the leftmost vertex being more acute than the corresponding parts
of the Fragment A daleths. However, the differences between the forms in
the respective fragments are not so great as to necessitate the view of two
mutually exclusive inscriptions. The variations in form within Fragment A
alone are enough to demonstrate that the style of daleth in Fragment A
could easily accommodate the style of daleth in Fragment B. A com-
parison between daleth A-2 and daleth B-2 reveals how close their respec-
tive forms are. A survey of the daleths in the Amman Citadel Inscription
also demonstrates the use of similarly divergent forms.
Another noteworthy point is the slant of the letter. A distinct leftward
slant is observable for the daleths of Fragment B. Consequently, the form
is at least contemporary with the form of Fragment A. This being the case,
we will have a similar register of matches with other Northwest Semitic
inscriptions. As in Fragment A, the form here displays distinctly Syrian
characteristics, which give us a terminus a quo in c. 825 BCE with the Kila-
muwa Inscription and the Amman Citadel Inscription. The only other dis-
tinguished matches are the bricks from Hamath, the stele of Bar Rakib, the
Siloam Tunnel Inscription, and even the ceramic texts of the Samaria
Ostraca.
With this information, we may place the daleth of Fragment B between
the late ninth century BCE and the end of the eighth century BCE, with the
form attested in Southeast Anatolia, Syria, Palestine and Transjordan. This
gives us the exact same bracket as that of the daleth in Fragment A. The
minor differences in the forms between Fragment A and Fragment B are
noteworthy, but are not conclusive for determining the relationship (or
lack of such) between the two fragments.
Heh. There are five instances of the letter heh in Fragments Bl and B2.
Table 4.24 contains their images in order of their appearance.
The form consists of a leftward leaning stem from which three 'bristles'
protrude to the left. These 'bristles' hang down heavily so that there is an
acute angle between the 'bristles' and the stem. The 'bristles' also appear
cluttered very close to each other. From the three fully preserved figures
4. Palaeographical Analysis 143
we observe that the bottom 'bristle' branches from the approximate mid-
point of the stem. Likewise, on these three particular figures we observe
that the middle 'bristle' is shorter than the other two.
doubts about any relationship between them. Indeed, the angles of some
strokes and the slant of some entire figures in single texts among the Sefire
Treaties are vastly more disparate than the differences between Fragment
A and Fragment B of Tel Dan. One of the consistencies between the two
fragments is the notable slant of the figures.
The second difference noted between the fragments was that the 'bris-
tles' in Fragment A tended to become shorter further down the stem. How-
ever, this was detected in only three of the six figures ofhehin Fragment
A (hehs A-2, A-4 and A-5). The other three figures displayed 'bristles' of
equal length. Thus, even to begin with, this is not a universal trend. Fur-
thermore, heh B-5, despite damage to the lower half of the letter, displays
the top and middle 'bristles' of the figure as being of equal length. Thus,
even within Fragment B itself, the tendency to have a shorter middle 'bris-
tle' is not universal. Thus, heh B-5 is for all intents and purposes, identical
to one of the more equal 'bristled' hehs on Fragment A, such as heh A-6.
Also, a survey of forms within other individual inscriptions reveals even
wider discrepancies than those represented by the Tel Dan fragments.
Again, the stelae of the Sefire Treaties contain a wealth of examples.
Another inscription that demonstrates this coexistence of seemingly diver-
gent forms is the Eqron Inscription. There we see figures with a short mid-
dle 'bristle' (e.g. line 3), as well as figures with 'bristles' diminishing in
size (e.g. line 5). Therefore, the difference in the length of the 'bristles'
between Fragment A and Fragment B are certainly not evidence of dif-
ferent hands.
The third difference between the hehs of Fragment A and Fragment B
which we noted was the length of the stem. Again, however, a survey of
other Northwest Semitic inscriptions shows this to be a misleading differ-
ence. First, a measurement of the absolute length of the stems in each
fragment shows them to be practically identical. Thus the difference in
stem length is merely an allusion created by the cluttering of the 'bristles'
in the hehs of Fragment B. Yet, in case there is any doubt, another survey
of Northwest Semitic inscriptions shows even wider varying stem lengths
within the one inscription.
Thus, when Fragment A and Fragment B are compared with other North-
west Semitic inscriptions, the difference in forms is seen to be miniscule.
In fact, the figures are shown to be remarkably similar to each other. The
differences that do exist may, therefore, be attributed to the natural inabil-
ity of one human hand to reproduce more than one written form in exact
replica. Yet, it must be said that few other individual inscriptions reproduce
4. Palaeographical Analysis 145
such closely matching figures. Therefore, the letter heh lends support to
the notion that Fragment A and Fragment B are indeed part of the one in-
scription, or were at least written by the same hand.
This inevitably means that we will have the same chronological horizon
for the hehs of Fragment B as we had for those of Fragment A. A date for
the form between the mid-ninth century BCE and the early seventh century
BCE is appropriate. Among the matches are the Melqart Stele, Hazael's
Horse Ornament, Kilamuwa, the Amman Citadel Inscription, the Nora
Stone, the Archaic Phoenician Inscription from Cyprus, Zakkur, Panammu
I, the Sefire Treaties, Bar Rakib, the Silwan Tomb, the Nerab Stelae and
the Eqron Inscription. This sees the form attested in Sardinia, Cyprus,
Southeast Anatolia, Syria, Palestine and Transjordan.
Waw. There are three instances of the letter waw in Fragments B1 and B2.
Table 4.25 contains the images of these in their sequential order.
The form consists of a vertical stem from which a 'hook' branches off to
the left. The stem has a discernable kink at this junction so that the top
portion of the stem bends very slightly to the right. In waw B-1 the 'hook'
has the appearance of being round. However, a definite vertex is discerna-
ble, meaning the 'hook' is made up of two strokes as in Fragment A. This
is also confirmed by the two observable deep points of chisel markings
inside the actual incisions.
Waw B-3, however, lacks the vertical stroke of the hook. This peculiar-
ity was understood by Cryer as proof that Fragment A and Fragment B are
from separate inscriptions, with the comment that it is only with great dif-
ficulty that this particular waw (B-3) can be understood as a waw at all.92
However, there is a good reason why this particular waw is missing half of
93. For this reason, the pockmark is included in the image of the figure in Table
4.25.
94. Cryer, 'King Hadad', p. 225.
95. B. Becking, 'The Second Danite Inscription: Some Remarks', BN81 (1996),
pp. 21-29(22).
96. Cryer has since altered his opinion on these issues in preference for the view
that all the fragments are forgeries.
97. Becking, 'The Second Danite Inscription', p. 22.
4. Palaeographical Analysis 147
The figure is similar to a Latin 'H' fallen on its side. It consists of two
parallel strokes that are joined by a connector stroke. The parallel strokes
are not quite horizontal as they slant slightly downwards on the left, thus
giving the overall impression that the letter leans towards the left. Simi-
larly, the connector stroke is not exactly vertical but leans to the right.
Both parallel strokes are of equal length.
Since there is no attestation of the letter zayin in Fragment A, we cannot
offer comparison with the form here in order to determine any relationship
between the two fragments. We can, however, find a niche for the form in
the chronology of script development among Northwest Semitic inscrip-
tions.
The particular form of zayin used in Fragment B represents an early stage
in the development of the script. A cursive style of zayin, resembling a Latin
'Z', becomes the predominant form at the end of the eighth century BCE.
On the Ahiram Sarcophagus, we find a precedent for the form attested
here in Fragment B. Two key differences may be noted, though. First, the
148 The Tel Dan Inscription
zayin of Ahiram has relatively short parallel strokes compared to the zayin
of Fragment B. Second, the figure in Ahiram characteristically leans to the
right. This is typical of an early style, evidenced by other Phoenician in-
scriptions such as the Shipitba'al Inscription. Later styles reflect a general
lengthening of the parallel strokes and a slight shift in rotation so that the
figures appear to sit flat or lean leftwards.
The first match comes with the Gezer Calendar. The difficulty with it,
however, is that the parallel strokes are on an inordinately steep slant so
that the figure appears grossly overweighted on the left. Also, the multi-
divergent forms of all letters represented in the Gezer Calendar means we
can by no means be sure that this single attestation of zayin in the inscrip-
tion is reliable as a measure of other forms. For these reasons, we must
refrain from attributing much weight to the match.
A meaningful match is made with the Melqart Stele, in which we see a
form comparably slanted to the one in Fragment B. The only difference is
the angle of the connector stroke. On the Melqart Stele, this stroke leans to
the left whereas in Fragment B it leans to the right. Only on one occasion
does the connector stroke lean to the right on the Melqart Stele (line 4,
second letter). Yet this form is closer in configuration to the cursive style
of zayin.
The angle of the connector stroke is the sole factor that prevents a
complete match with any dedicatory inscription from the Northwest
Semitic corpus. Except for the questionable figure in the Gezer Calendar,
and the solitary exception mentioned in the Melqart Stele, all non-cursive
styles of zayin display the connector stroke as perpendicular to the other
two strokes. This configuration is seen not only on the Melqart Stele, but
also on Kilamuwa's Sceptre, in Panammu I and the Sefire Treaties. In the
Mesha Stele, Kilamuwa Stele and the Azitawadda Inscription, we see
similarly configured forms that sit quite flat on the writing base line.
Except for this factor, the forms are comparable to that of Fragment B. It
is unfortunate that the letter zayin is not attested on either the Nora Stone
or the Amman Citadel Inscription.
However, with this information, we can make some important observa-
tions about the form in Fragment B and so make good inroads towards a
palaeographic dating of the letter. First, all non-cursive styles of zayin put
the connector stroke perpendicular to the other two parallel strokes. The
only exception is the odd form in the Gezer Calendar and one form from
the Melqart Stele. It is clear from the connector stroke of the Fragment B
4. Palaeographical Analysis 149
zayin that the figure is intended as a non-cursive form. The fact that the
connector stroke touches the bottom stroke at the middle demonstrates this
well. However, the angle is far from perpendicular to the top and bottom
strokes. In fact, the angle of the stroke is 45 to the top and bottom strokes.
This suggests that the form here in Fragment B is some way between the
non-cursive and the cursive styles of zayin.
Unfortunately, there is no way of proving this suggestion because we
have only one example with which to work. The single cursive-like figure
on the Melqart Stele shows that it need not be a universal style within an
inscription either. The earliest cursive style of zayin we have is Hazael's
Horse Ornament from Samos from the second half of the ninth century
BCE. The latest non-cursive style zayin on a dedicatory inscription is from
the Siloam Tunnel Inscription and the Silwan Tomb at the end of the
eighth century BCE. Thus, I proffer this period of overlap between the two
styles, from the mid-ninth century BCE to the end of the eighth century
BCE, as a reasonable bracket in which to place the style of zayin.
The zayin of Fragment B is, therefore, seen to be a very rare form of the
letter. The closest comparable forms come from Southeast Anatolia, Syria,
Palestine and Transjordan.
The form of the heth here in Fragment B is very similar to that encoun-
tered in Fragment A. If anything, the proportion of the 'risers' here in
Fragment B seems to be slightly more exaggerated than those in Fragment
A. However, we have the same style in both forms, demonstrated primar-
ily by the 'risers'. First, the left 'riser' extends notably above the top
'rung' and only slightly below the bottom 'rung'. For the right 'riser' we
see the opposite, in which it protrudes only minutely above the top 'rung',
but slightly more below the bottom 'rung'.
Cryer's description of this heth is not accurate. In comparing the two
forms from both fragments, he states that the form in Fragment A is
150 The Tel Dan Inscription
unusual in displaying the left 'riser' higher than the right 'riser'.98 In
making the comparison, he states that such is not the case for the form in
Fragment B. Cryer's statement is somewhat ambiguous in that he does not
clarify whether he is referring to the protrusion of the 'risers' beyond the
'rungs', or to the overall latitude at which each 'riser' is placed on the
stone. In either case, Cryer's accusation carries no weight. If he means the
former, then he is clearly mistaken as to the configuration of the letter in
Fragment B, which is the same as that in Fragment A. If he means the
latter, then it is of no consequence, for the difference in latitude is merely
a question of the figures' respective rotations. The figure in Fragment B
leans slightly more leftward than the figure in Fragment A. Yet, both fig-
ures display clear rotation. Cryer's influence on Becking in this regard is
unfortunate for both use this inaccurate observation as an argument for
two different styles ofheth and, therefore, two separate inscriptions." The
truth is quite the opposite as the forms show a very close association with
each other and provide good support for the fragments being from the
same inscription.
This being the case, the matches made for the form in Fragment A will
apply for the form here also. Matches can be made with Tell Fakhariyah,
the Hazael Ivory, Kilamuwa, the Amman Citadel Inscription (even though
it has only two 'rungs'), Zakkur and the Siloam Tunnel. This yields a
chronological bracket between the mid-ninth century BCE and the end of
the eighth century BCE, with the forms attested in Southeast Anatolia,
Syria, Palestine and Transjordan.
Yodh. There are four certain instances of the letter yodh in Fragment B.
The yodh proposed by Biran and Naveh at the beginning of Line 4 cannot
be verified and cannot, therefore, count in our palaeographical analysis.
The four certain instances of yodh and their images are contained in Table
4.28.
The forms of yodh here in Fragment B are similar to those found in
Fragment A. The two fully preserved figures (yodhs B-l and B-3) show
characteristics of both curved and angular style yodhs. This tendency was
also observed on Fragment A. Yodh B-l closely resembles yodh A-8 from
Fragment A and yodhs B-2 and B-3 show similarity with yodhs A-2, A-7
and A-9. Thus, there is good correspondence between the two fragments,
suggesting a connection between them.
B-3 Line 6 ro
Since there is such close correspondence between the forms, the same
palaeographic considerations for theyodhs of Fragment A inevitably apply
to the yodhs of Fragment B as well. Thus, matches for the forms can be
made with the Melqart Stele, the Amman Citadel Inscription, the Archaic
Phoenician Inscription from Cyprus, Kilamuwa, Zakkur, Panammu I, the
Phoenician Governor's Bowl from Cyprus, the Sefire Treaties, Bar Rakib,
Azitawadda, the Nimrud Ivory Plaque, the Nerab Stelae and the Eqron
Inscription. This gives us a chronological scope for the yodh between the
mid-ninth century BCE and the early seventh century BCE, with forms from
Cyprus, Southeast Anatolia, Syria, Palestine and Transjordan.
Kaph. There are two instances of the letter kaph in Fragment B, both of
them occurring in Fragment B1. Their images are seen below in Table 4.29.
100. The leftmost portions of this figure have been destroyed in breakage.
101. The bottom portions of this figure have been destroyed in breakage.
102. The leftmost portions of this figure have been destroyed in breakage.
152 The Tel Dan Inscription
Lamed. There are four instances of the letter lamed in Fragment B, all of
them occurring on Fragment B1. Their images are contained in order of
their appearance in Table 4.30.
B-2 Line 3 ^1
104. These three colleagues were N. Cohen, A. Weiner and R. Yekutieli. See Biran
and Naveh, 'The Tel Dan Inscription', p. 11 n. 5.
105. The top portion of the 'shank' on this lamedhas been destroyed by breakage.
154 The Tel Dan Inscription
All of the figures are in the shape of a 'fish hook'. The 'shanks' show little
or no curvature, while the actual hook is a neat curve. There is evidence of
a slight 'bump' in the 'hook' oflamedB-2. The 'barbs' on all figures slant
noticeably upwards, pointing essentially in the same direction as the top of
the 'shank'.
Becking notices a difference between the form here in Fragment B and
the form represented in Fragment A.106 He does not, however, specify the
nature of the difference. We can only surmise that it is the uniform curva-
ture of the 'hook' that Becking has in mind. The figures attested on Frag-
ment A do not display such uniform roundness at the 'hook', but are prone
to hyperbolic roundness which makes the 'hook' look like a distinct bend
or 'bump' rather than a smooth curve.
However, lamedB-2 also shows evidence of this 'bump' and is, in fact,
very similar to lamed A-4 on Fragment A. In addition to this, lamed A-9
shows uniform curvature of the 'hook' so that it is similar to the forms on
Fragment B. Thus, there is adequate overlap of the forms on both frag-
ments, and so the figures do not lend support to Becking's claim of exclu-
sive differences.
Since there is such considerable overlap between the two forms, the
matches for the lamed of Fragment B significantly overlap the matches for
the form in Fragment A. Our earliest matches are with the Phoenician
inscriptions of Ahiram, Yehimilk and Shipitba'al. The Mesha Stele forms
are comparable in all aspects except the 'barbs', which tend to hook more
to the vertical. The same may be said for the El-Kerak Fragment. No such
difficulties lie with the forms from Kilamuwa, the Amman Citadel Inscrip-
tion, the Archaic Phoenician Inscription from Cyprus, Zakkur, Panammu I,
the Sefire Treaties, the Hamath Bricks, Azitawadda, Panammu II, Bar
Rakib, the Nerab Stelae and the Tema Stele. As was also the case with the
form of Fragment A, numerous smaller texts from Palestine, such as seals,
ostraca and the like, testify to the persistence of the form right through to
the mid-fourth century bulla of Yesha'yahu ben-Sanballat, Governor of
Samaria.
Thus, the data on the form of lamed gives us a huge chronological
horizon stretching six centuriesfrom the mid-tenth century BCE through
to the mid-fourth century BCE. The form is attested throughout the entire
Levant as well as Cyprus and Sardinia.
Mem. There are five instances of the letter mem in Fragment B. Their
images are contained in order of their appearance in Table 4.31.
B-4 Line 5 fa
There is a slight difference in the forms here in Fragment B. The first three
figures (mems B-l, B-2, B-3) have a continuous 'jagged head' in which
each of the four strokes matches up with the end of another stroke. On the
last two attestations (mems B-4 and B-5), the two centre strokes of the
'jagged head' do not meet end to end. Rather, it appears that the left two
strokes of the 'jagged head' have slipped down slightly.
This internal inconsistency in the form of mem, however, does not affect
the correspondence in the forms with Fragment A. On the contrary, it does
much to enhance it. For the mem of Fragment A, we noticed that the char-
acteristic markings of the form were the angle of the leftmost stroke of the
'jagged head' and the angle of the 'jagged head' as a whole. When we
compare these points with those of the forms here in Fragment B, we ob-
serve an identical match. Furthermore, the curvature of the stem matches
in every way. This means that the only difference between the two forms
of mem is the slight disconnection of the 'jagged head' seen here in mems
B-4 and B-5. Yet, this is an irrelevant difference because it is internal to
Fragment B and the first three figures there show close correspondence
with the forms in Fragment A.
The attestation of such a slight shift within the form of mem is a crucial
datum for the relationship between Fragment A and Fragment B. This will
156 The Tel Dan Inscription
Nun. There are two instances of the letter nun in Fragment B. Both of
these, however, are severely damaged. The visible parts of these figures
are contained in Table 4.32.
107. Since this letter is mostly damaged by a lacuna, the image is not very clear.
4. Palaeographical Analysis 157
Semitic inscriptions is also difficult. We may only note that there appear to
be similarities with the forms of Fragment A, and therefore we suspect the
same matches with other inscriptions as those made by Fragment A. Com-
prehensive comparison, however, cannot be made.
'Ayin. There are three instances of the letter 'ayin in Fragment B. Their
images are seen below in Table 4.34.
As in Fragment A, the letter is in the shape of a simple closed circle.
The two notable characteristics of the form from Fragment A are also
matched herenamely, the lack of a pictographic dot in the centre of the
figure and the overall size of the letter. In addition, we note the
characteristic 'bump' at the bottom left of 'ayin B-l. Although this is not
108. Since this letter is mostly damaged by a lacuna, the image is not very clear.
158 The Tel Dan Inscription
visible on 'ayin B-2, the curvature at the bottom of the letter hints at such
a 'bump'.
109. The left portion of this 'ayin has been destroyed by breakage.
110. This figure is mostly destroyed due to the breakage of the stone. In my epi-
graphical analysis, I noted that there was the possibility this character could be restored
as something other than 'ayin, but that we should restore this letter as an 'ayin on phi-
lological grounds.
111. Cryer, 'KingHadad', p. 226.
4. Palaeographical Analysis 159
Peh. There is a single occurrence of the letter peh in Fragment B. Its image
is shown below in Table 4.35.
The figure is a fluid curved stroke and is essentially the same as the form
used in Fragment A. The form here in Fragment B is slightly longer than
the form used in Fragment A, but certainly not long enough to demand a
different hand. On the contrary, the fluidity of the figure closely resembles
peh A-l, suggesting a connection between the two forms.
Since there is such a close correlation between the forms of both frag-
ments, we also see a correlation with inscriptions containing comparable
forms of peh. Thus, we find good matches with the Archaic Phoenician
Inscription from Cyprus, and with the Nora Stone. On the Levantine main-
land, matches are found in the Zakkur Stele, the Sefire Treaties, Azitawad-
da, Panammu II, Bar Rakib and the sarcophagi of Tabnit and Eshmun'azar.
These data provide us a chronological bracket identical to the form in
Fragment Anamely, between the late ninth century BCE and the early
fifth century BCE. The form is attested in Sardinia, Cyprus, Southeast
Anatolia, Syria and Phoenicia.
Qoph. There is one instance of the letter qoph in Fragment B. Its image is
shown below in Table 4.36.
The form is identical to that encountered in Fragment A. As such, all
considerations for the form in Fragment A apply to this form also. They
match in shape, proportion and slant. This close correlation between the
two forms certainly suggests that the same hand lay behind both fragments.
160 The Tel Dan Inscription
Once again, the fine correspondence between the forms means our
matches for the form in Fragment A are the same for Fragment B. There-
fore, our matches for Fragment B range from the Melqart Stele in the mid-
ninth century BCE down to the Eqron Inscription in the early seventh
century BCE. Between these dates, we find matches with Hazael's Horse
Ornament, both Kilamuwa's Stele and Sceptre, the Archaic Phoenician In-
scription from Cyprus, Zakkur, Panammu I, the Sefire Treaties, Azita-
wadda, Panammu II and the Nerab Stelae.
Therefore, our chronological spectrum for the qoph of Fragment B, like
that of Fragment A, ranges from the mid-ninth century BCE to the early
seventh century BCE, with the form attested in Cyprus, Southeast Anatolia,
Syria and Palestine.
Resh. There are five instances of the letter resh on Fragment B. Their
images are contained in order of their appearance in Table 4.37 below.
Table 4.37. Occurrences o/resh in Fragment B.
No. Image Location Occurring Word
B-4 Line 7 in
112. This letter has been restored as a resh on both epigraphical and philological
grounds. The 'head' of the figure has been destroyed in breakage.
4. Palaeographical Analysis 161
lesser degree, Bar Rakib. The sarcophagi of Tabnit and Eshmun'azar also
qualify as matches.
Thus, the chronological bracket for the form of resh in Fragment B is
between the mid-ninth century BCE and the mid-fifth century BCE, with the
form attested in Cyprus, Southeast Anatolia, Syria and Transjordan. The
fact that many of the matching inscriptions are the same as those found for
Fragment A indicates that both forms of resh attested on the Tel Dan frag-
ments are not necessarily mutually exclusive. Rather, they appear to be
variations on the same form. This further undermines the objections raised
by Cryer and Becking in relation to this letter.
115. Since this sin is right on the join between Fragments Bl and B2, the image is
not very clear.
4. Palaeographical Analysis 163
All that remains of this taw is the bottom portion of the stem and the
leftmost point of the crossbar. Despite this, we may still compare it with
the relevant portions from the form of taw in Fragment A. The difference
is shown to be the curvature of the stem. In Fragment A, the stem of the
taw did not have such a pronounced curl. However, three of the figures
(taws A-l, A-3 and A-5) have a very modest flourish at the bottom of the
stem. In this way, the remnant of the taw in Fragment B is not too far
removed from the relevant portions of the figures in Fragment A.
Since the top half of the figure is missing, it is meaningless to pursue
matches for the taw of Fragment B with other inscriptions. We can, how-
ever, note that the form is a long-stemmed taw, rather than the shorter equi-
lateral taw. This shows a clear Syrian influence over the form of the letter.
116. Only the bottom portion of the stem and the leftmost point of the cross-bar have
been preserved of this taw. The rest of the figure has been broken off.
164 The Tel Dan Inscription
This being the case, we see the Amman Citadel Inscription and the Zakkur
Stele as the closest matches, just as in Fragment A. These two particular
inscriptions are separated by less than half a century around the turn of the
eighth century BCE. Since heth is one of the most complex letters, it is a
fairly accurate indicator of variant styles.
The forms of 'aleph and 'ayin employed in Fragment B point to a date
slightly after 825 BCE. At that time, the Kilamuwa Stele still used early
styles of these letters. Thus, we find ourselves in the same time period as
that assigned to Fragment Anamely, c. 800 BCE ( 20 years).
Second, the form of the letter mem was seen to be inconsistent within
Fragment B itself. One form was practically identical to that encountered
in Fragment A, while the other represented a form in which half of the
'jagged head' had slipped slightly downwards. This internal inconsistency
alerts us to the tendency of the scribe of Fragment B to make slight alter-
ations to particular letters. This visible trend can be legitimately applied to
other letters which vary slightly between the fragments, such as the
slightly wider daleth, the cluttered 'bristles' on the heh and the curling at
the bottom of some long-stemmed letters in Fragment B. With this scribal
tendency, a firm connection between Fragment A and Fragment B is estab-
lished. This accords with the data about the engraving of the individual
letters, which was seen to be identical in both Fragment A and Fragment
B. We may thus talk of the fragments in the context of one inscription,
which we will henceforth term, the Tel Dan Inscription. My palaeographi-
cal analysis also confirms the veracity of the date for the fragments yielded
in our examination of the archaeological context. As such, we may legiti-
mately claim that the Tel Dan Inscription was written at some time close
to 800 BCE.
i rwYvjivij^i-N i .r^.
FRAGMENT A
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
3 Bar Rakib (cont.)
(cont.) Panammu II
Eqron
Tabnit Sarcophagus
Eshmun'azar Sarcophagus
Tema
3
1 Kilamuwa Palestine Late-ninth-
Amman Citadel Southeast Anatolia late eighth
(Samaria Ostraca) Syria
Hamath Bricks Transjordan
Bar Rakib
Siloam Tunnel
71 Melqart Stele Cyprus Mid-ninth-
Mesha Stele Southeast Anatolia early seventh
Amman Citadel Syria
Archaic Phoenician (Cyprus) Transjordan
Nora Inscription
Hazael's Horse Ornament
Zakkur
Bar Rakib
Nerab Stelae
Eqron
1 Kilamuwa Cyprus Late ninth-
Archaic Phoenician (Cyprus) Southeast Anatolia late-eighth
Zakkur
Panammu I
Sefire Treaties
Panammu II
Bar Rakib
T
n Tell Fakhariyah Palestine Mid-ninth-
Hazael Ivory Samal late eighth
Kilamuwa Syria
Amman Citadel Transjordan
Zakkur
Siloam Tunnel
C3
Melqart Stele Cyprus Mid-ninth-
Kilamuwa Palestine early seventh
Amman Citadel Southeast Anatolia
Archaic Phoenician (Cyprus) Syria
Zakkur Transjordan
Panammu I
Phoenician Governor's Bowl
Sefire Treaties
Azitawadda
4. Palaeographical Analysis 167
FRAGMENT A
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
* Bar Rakib
(cont.) Nimrud Ivory Plaque
Nerab Stelae
Eqron
D Mesha Stele Cyprus Late ninth-
Kuntilet 'Ajrud Negev early seventh
Amman Citadel Palestine
Nora Inscription SE Anatolia
Phoenician Governor's Bowl Sardinia
Panammu I Syria
Hamath Bricks Transjordan
Bar Rakib
Siloam Tunnel
Eqron
Ahiram Sarcophagus Cyprus Mid-tenth-
^ Yehimilk Negev mid-fourth
Shipitba'al Northwest Arabia
Mesha Stele Palestine
El-Kerak Fragment Phoenicia
Kilamuwa Southeast Anatolia
Amman Citadel Sardinia
Archaic Phoenician (Cyprus) Syria
Nora Inscription Transjordan
Kuntilet 'Ajrud Wares
Zakkur
(Samaria Ostraca)
(Samaria Seals)
Panammu I
Sefire Treaties
Hamath Bricks
(Khirbet el-Qom)
(Ophel Ostracon)
Azitawadda
Panammu II
Bar Rakib
(Tel Qasile Ostracon II)
(Beersheba Juglet)
Nerab Stelae
(Lachish III Seals)
(Ramat Rahel Seal)
(Tell en-Nasbeh Seal)
(Wadi Muraba'at Papyrus)
(Shechem III Seal)
Tema
(Yesha'yahu Bulla)
Q Mesha Stele Southeast Anatolia Mid-ninth-
El-Kerak Fragment Syria late eighth
168 The Tel Dan Inscription
FRAGMENT A
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
a Kilamuwa Transjordan
(cont.) Amman Citadel
Zakkur
Panammu I
Sefire Treaties
Hamath Bricks
Panammu II
Bar Rakib
] Melqart Stele Samal Mid-ninth-
Mesha Stele Syria late eighth
Kilamuwa Transjordan
Amman Citadel
Zakkur
Panammu I
Sefire Treaties
Hamath Bricks
Panammu II
D Kilamuwa Scepter Southeast Anatolia Late ninth-
Amman Citadel Syria late eighth
Hamath Bricks Transjordan
Azitawadda
i; Mesha Stele Cyprus Mid-ninth-
Amman Citadel Palestine early fifth
Archaic Phoenician (Cyprus) Phoenicia
Panammu I Southeast Anatolia
Panammu II Syria
Nerab Stelae Transjordan
Siloam Tunnel
Silwan Tomb
Eqron
Tabnit Sarcophagus
Eshmun'azar Sarcophagus
s Archaic Phoenician (Cyprus) Cyprus Late ninth-
Nora Inscription Phoenicia early fifth
Zakkur Southeast Anatolia
Sefire Treaties Sardinia
Azitawadda Syria
Panammu II
Bar Rakib
Tabnit Sarcophagus
Eshmun'azar Sarcophagus
x Comparison not practicable
P Melqart Stele Cyprus Mid-ninth-
Hazael's Horse Ornament Palestine early seventh
Kilamuwa Southeast Anatolia
Kilamuwa's Sceptre Syria
Archaic Phoenician (Cyprus)
4. Palaeographical Analysis 169
FRAGMENT A
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
P Zakkur
(cont.) Panammu I
Sefire Treaties
Azitawadda
Panammu II
Nerab Stelae
Eqron
i Melqart Stele Cyprus Mid-ninth
Mesha Stele Northwest Arabia mid-fifth
Kilamuwa Southeast Anatolia
Kilamuwa Sceptre Syria
Amman Citadel Transjordan
Archaic Phoenician (Cyprus)
Panammu I
Sefire Treaties
Panammu II
Bar Rakib
Tabnit Sarcophagus
Eshmun'azar Sarcophagus
Tema
E? Tell Fakhariyah Cyprus Mid-ninth-
Amman Citadel Mesopotamia early seventh
Nora Palestine
Zakkur Southeast Anatolia
Panammu I Sardinia
Phoenician Governor's Bowl Syria
Sefire Treaties Transjordan
Azitawadda
Panammu II
Bar Rakib
Nerab Stelae
Eqron
n Melqart Stele Cyprus Mid-ninth-
Hazael Ivory Southeast Anatolia late eighth
Hazael's Horse Ornament Syria
Archaic Phoenician (Cyprus)
Kilamuwa
Zakkur
Panammu I
Sefire Treaties
Hamath Bricks
Azitawadda
Panammu II
Bar Rakib
Nerab Stelae
Date for script of Fragment A: c. 800 BCE ( 20 years)
170 The Tel Dan Inscription
FRAGMENT B
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
n Archaic Phoenician (Cyprus) Transjordan
(cont.) Zakkur
Panammu I
Sefire Treaties
Bar Rakib
Silwan Tomb
Nerab Stelae
Eqron
1 Kilamuwa Cyprus Late ninth-
Archaic Phoenician (Cyprus) Southeast Anatolia late eighth
Zakkur Syria
Panammu I
Sefire Treaties
Panammu II
Bar Rakib
T Melqart Stele Palestine Mid-ninth-
Kilamuwa Southeast Anatolia late eighth
Kilamuwa' s Sceptre Syria
Panammu I Transjordan
Azitawadda
Sefire Treaties
Siloam Tunnel
Silwan Tomb
n Tell Fakhariyah Palestine Mid-ninth-
Hazael Ivory Southeast Anatolia late eighth
Kilamuwa Syria
Amman Citadel Transjordan
Zakkur
Siloam Tunnel
B1
Melqart Stele Cyprus Mid-ninth-
Kilamuwa Palestine early seventh
Amman Citadel Inscription Southeast Anatolia
Archaic Phoenician (Cyprus) Syria
Zakkur Transjordan
Panammu I
Phoenician Governor's Bowl
Sefire Treaties
Bar Rakib
Azitawadda
Nimrud Ivory Plaque
Nerab Stelae
Eqron
D Mesha Stele Negev Mid-ninth-
El-Kerak Fragment Palestine early seventh
Kilamuwa Southeast Anatolia
Amman Citadel Inscription Syria
172 The Tel Dan Inscription
FRAGMENT B
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
D Kuntilet 'Ajrud Transjordan
(cont.) Nimrud Ivory Plaque
Azitawadda
Siloam Tunnel
Eqron
? Ahiram Sarcophagus Cyprus Mid-tenth-
Yehimilk Negev mid-fourth
Shipitba'al Northwest Arabia
Mesha Stele Palestine
El-Kerak Fragment Phoenicia
Kilamuwa Southeast Anatolia
Amman Citadel Sardinia
Archaic Cyprus Syria
Nora Inscription Transjordan
Kuntilet 'Ajrud Wares
Zakkur
(Samaria Ostraca)
(Samaria Seals)
Panammu I
Sefire Treaties
Hamath Bricks
(Khirbet el-Qom)
(Ophel Ostracon)
Azitawadda
Panammu II
Bar Rakib
(Tel Qasile Ostracon II)
(Beersheba Juglet)
Nerab Stelae
(Lachish III Seals)
(Ramat Rahel Seal)
(Tell en-Nasbeh Seal)
(Wadi Muraba'at Papyrus)
(Shechem III Seal)
Tema
Yesha'yahu Bulla
D Mesha Stele Southeast Anatolia Mid-ninth-
El-Kerak Fragment Syria late eighth
Kilamuwa Transjordan
Amman Citadel
Zakkur
Panammu I
Sefire Treaties
Hamath Bricks
Panammu II
Bar Rakib
] Comparison not practicable
4. Palaeographical Analysis 173
FRAGMENT B
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
0 Kilamuwa Sceptre Southeast Anatolia Late ninth-
Amman Citadel Syria late eighth
Hamath Bricks Transjordan
Azitawadda
y Mesha Stele Cyprus Mid-ninth-
Amman Citadel Palestine early fifth
Archaic Phoenician (Cyprus) Phoenicia
Panammu I Southeast Anatolia
Panammu II Syria
Nerab Stelae Transjordan
Siloam Tunnel
Silwan Tomb
Eqron
Tabnit Sarcophagus
Eshmun'azar Sarcophagus
s Archaic Phoenician (Cyprus) Cyprus Late ninth-
Nora Inscription Phoenicia early fifth
Zakkur Southeast Anatolia
Sefire Treaties Sardinia
Azitawadda Syria
Panammu II
Bar Rakib
Tabnit Sarcophagus
Eshmun'azar Sarcophagus
x
P Melqart Stele Cyprus Mid-ninth-
Hazael's Horse Ornament Palestine early seventh
Kilamuwa Southeast Anatolia
Kilamuwa Sceptre Syria
Archaic Phoenician (Cyprus)
Zakkur
Panammu I
Sefire Treaties
Azitawadda
Panammu II
Nerab Stelae
Eqron
n Melqart Stele Cyprus Mid-ninth-
Mesha Stele Phoenicia mid-fifth
Kilamuwa Southeast Anatolia
Kilamuwa Sceptre Syria
Amman Citadel Transjordan
Archaic Phoenician (Cyprus)
Panammu I
Sefire Treaties
Hamath Bricks
Panammu II
174 The Tel Dan Inscription
FRAGMENT B
Letter Closest Matching Inscriptions Regional Bracket Chronological Bracket
(centuries BCE)
~1 Bar Rakib
(cont.) Tabnit Sarcophagus
Eshmun'a/ar Sarcophagus
to Tell Fakhariyah Cyprus Mid-ninth-
Amman Citadel Mesopotamia early seventh
Nora Inscription Palestine
Zakkur Southeast Anatolia
Panammu I Sardinia
Phoenician Governor's Bowl Syria
Sefire Treaties Transjordan
Azitawadda
Panammu II
Bar Rakib
Nerab Stelae
Eqron
n Comparison not practicable Syrian Influence
Date for script of Fragment B: c. 800 BCE ( 20 years)
Chapter 5
Figure 5.1. The arrangement of the fragments proposed by Biran and Naveh with
Fragment B to the left of Fragment A (Tel Dan Excavations, Hebrew Union College,
Jerusalem; photograph: Z. Radovan).
Epigmphical Considerations
Doubts about the integrity of the original arrangement were first raised by
Cryer and Thompson of the University of Copenhagen in two articles pub-
lished in the Scandinavian Journal of the Old Testament.1 This occurred
immediately following the publication of Fragment B in the Israel Explo-
ration Journal by Biran and Naveh. Cryer's objection to the arrangement,
indeed the entire association of the fragments, was based primarily on
palaeographic considerations. He saw the differences between certain
letterforms as evidence of the mutual exclusivity of the fragments.
Thompson's objections, which led him to the same conclusion, were
based primarily on archaeological and epigraphic grounds. Thompson ques-
tioned whether Fragment A and Fragment B had a common archaeological
context seeing as they were discovered in different usages at different loca-
tions in the Israelite gate complex at Dan. Epigraphically, Thompson
noticed that most of the lines of text in the original arrangement did not
properly align between Fragment A and Fragment B. He also stated that
there were insurmountable problems regarding the room required to accom-
modate certain letters proposed by Biran and Naveh between Fragment A
and Fragment B.
I will deal with Thompson's epigraphical arguments here. First, his
objection to the alignment of the textual lines between Fragment A and
Fragment B is quite cogent. The arrangement proposed by Biran and
Naveh et al. aligns the fragments at Line 5. In so doing, however, there are
great discrepancies in the alignment of other lines (see Fig. 5.2). It may be
claimed that this misalignment is simply due to the lack of ruled guiding
lines to aid the scribe writing the text. However, it was demonstrated that
the trends of the lines in Fragment A sloped downwards because the scribe
had to stretch forward in order to write. This would not have altered for
Figure 5.2. The incongruence of line trends between Fragment A (right) and Fragments
(left) (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan
[detail]).
Schniedewind understood the problems with both the physical join and
the alignment of respective lines. He noted the possibility that there may be
a larger gap between Fragment A and Fragment B.8 This, however, creates
a series of other problems, most notably textual. Schniedewind was able to
overcome this difficulty by simply proposing a rotation of the two B frag-
ments. He rotated Fragment B1 2 in an anti-clockwise direction and Frag-
ment B2 1.5 in the same direction. Thus, Fragments B1 and B2 were offset
by 0.5. According to Schniedewind, this 'slight rotation of the fragments
makes the lines match better and renders a more convincing join'.9
However, numerous objections must be raised to Schniedewind's pro-
posal of fragment rotation. First, Schniedewind was working only with
computer imaging of the fragments. He did not test his hypothesis on the
actual fragments. To date, his proposal has still not been carried out on the
actual physical fragments. Thus, there is no way of knowing whether rota-
tion of the fragments is physically possible.
Nevertheless, Schniedewind's claim that the rotation produces a better
alignment and more convincing join between the fragments cannot be sub-
stantiated. Although rotation aids in the alignment of some lines above
Line 5, those below it suffer deterioration. Line 8 is particularly disparate.
Furthermore, the proposed join beneath the surface would be compromised
as there would be less contact between Fragment A and Fragment B. In
short, Schniedewind's proposal does not bring us any closer to substantiat-
ing the join proposed by Biran and Naveh et al.
It is clear from the trends of the lines that different physical circum-
stances governed the writing of Fragment A and Fragment B. The trend of
Fragment A was consistent with writing in the upper portions of the stone,
while the trend of Fragment B was shown to be consistent with writing at
the bottom of the stone. Therefore, the trends of the lines suggest placing
Fragment B beneath Fragment A, rather than to the left of it.
More weight is added to this suggestion when it is seen that many of the
longer characters encroach on the line beneath. For example, the mem and
kaph in Line B4 have stems that reach down to the same level as the top of
the letters in Line B5. The mem in particular nearly touches the 'head' of
the qoph beneath it. Similarly, the stem of the damaged nun in Line B5
almost touches the resh directly beneath it in Line B6.1 suspect that the
mem preceding this nun was similarly carved so that the stem almost
touched the samekh beneath it. I also note that the mem in Line B7 almost
joins with the waw beneath it in Line 8. These characteristics are con-
sistent with a scribe who reached the bottom portion of the stone and
needed to cram his letters slightly.
Palaeographical Considerations
The slight difference in the forms of some letters led Cryer to claim the
mutual exclusivity of Fragment A and Fragment B. Others, like Becking,10
soon followed with similar hypotheses. In the earlier palaeographical analy-
sis of the fragments, we saw how these differences were not sufficient to
warrant the interpretation that we were dealing with two separate inscrip-
tions. Rather, they were seen to be variant forms, many of which had forms
common to both Fragment A and Fragment B (e.g. lamed and mem).
However, in trying to account for why these variant forms exist at all,
we find further evidence for placing Fragment B below Fragment A, rather
than to its left. It was noted that letters in Fragment B were quite neat and,
when compared with the letters in Fragment A, they were neater on the
whole. That is, lines were straighter, curves were more rounded and small
flourishes were achieved. This suggests a greater ease of movement in the
arm and greater control of the writing implement. Thus, the scribe's arm
must have been in a very comfortable position during writing of the lines
in Fragment B.
If we keep the current arrangement proposed by Biran and Naveh, we
see a great disparity in the ease of movement between one side of the
inscription and the other. The first eight lines of Fragment A show signs of
strained movement whereas no such strain is visible for the writing of the
corresponding lines in Fragment B. Such longitudinal disparity is very
unlikely. In fact, if there was to be any longitudinal disparity in the ease of
movement, then we would expect more strain in writing to be evident in
the left side than the right, due to a crossing of the scribe's arm. Thus, the
arrangement of placing Fragment B to the left of Fragment A represents a
very unusual irregularity in the style of writing. Rather, the ease of move-
ment evident in the writing of Fragment B demonstrates that it should be
placed below Fragment A. Five particular letters highlight this for us.
Beth. The form ofbeth used in Fragment A was the less common curved-
stem beth. This was seen to be a variation of the more common vertexed-
stem beth employed in Fragment B. A bridge between the two forms was
seen in the figures of beths A-l and A-5. I propose that the curved-stem
beth was employed in Fragment A because of the scribe's strain in writing.
Due to the posture needed to write in the upper portions of the stone,
vertexes were hard to achieve. As such, instead of employing the more
common vertexed-stem beth in the upper portions, the scribe sufficed or
resorted to using the close variant of curved-stem beths. In the lower por-
tions of the stone, however, no such strain was exerted in writing. As such,
the scribe was able to write vertexed-stem beths with ease.
Heh. The difference between the hehs was seen to be the distance between
the 'bristles' as well as their lengths. The form in Fragment A puts a
noticeable gap between the 'bristles', which tend to become shorter further
down the stem. This suggests, once again, that the scribe experienced
some strain in writing Fragment A. This strain prevented him from being
able to accurately place the 'bristles' close together. Rather, with his arm
being stretched forward, the scribe would naturally have wanted to pull his
arm closer to himself for more comfort. This natural force of retracting the
arm led him to put some space between the 'bristles'. The stretched pos-
ture also prevented him from making the 'bristles' of comparable length.
Heh A-6, the last heh on Fragment A, shows signs of the scribe beginning
to experience greater control in writing. When compared with other hehs,
the 'bristles' of heh A-6 are closer together and are of comparable length.
Samekh. The same force at work in the letter heh applied to the writing
of the letter samekh. The difference in form between the fragments is the
spacing of the crossbars. In Fragment A, the crossbars are noticeably
spaced, whereas in Fragment B they are much closer together. Fragment A
shows signs of the scribe's arm wanting to retract from the strain of
stretching forward, hence the wide spacing. No such factor is evident in
the samekh of Fragment B, suggesting a greater ease of movement.
Resh. The same factor that was seen for the letter daleth applied to the
writing of the resh. That is, the 'head' of the resh in Fragment A is notice-
ably taller than the 'head' of the resh in Fragment B. This is consistent
with the scribe stretching forward and having considerably less ease in
writing on Fragment A than he did in writing on Fragment B.
Thus, the palaeographical evidence suggests that Fragment B should be
placed below Fragment A.
Textual Considerations
The main reason cited by Biran and Naveh for their particular arrangement
(see Fig. 5.3) of the fragments is the reading of the text that ensues. There-
5. Arrangement of the Fragments 183
fore, we must test to see whether their reading of the text is sustainable.
Numerous scholars have raised objections to the publishers' text.11 In-
deed, most of the controversy surrounding the Tel Dan Inscription has
been textual in nature. The single most important textual consideration is
the translation of the enigmatic word "TITf D in Fragment A, Line 9. How-
ever, there are other factors that hold the text of Biran and Naveh together.
Most of these have to do with the flow of text between the fragments. I
shall deal with these line by line. Since Biran and Naveh offer no connec-
tive text in Line 1,1 shall start with their reconstruction of Line 2.
11. E.g., and among others, Knauf, De Pury and Rmer; Garbini; Cryer; Thompson;
and Davies (see individual entries in the Bibliography).
184 The Tel Dan Inscription
word divider is the equivalent to half the space of a normal letter. In order
to accommodate the words required by Biran and Naveh, the letters would
have to be considerably crammed together so as to overlap each othera
suggestion which cannot really be sustained with any credibility. Thus, the
reconstructed text of Biran and Naveh is physically impossible here.
Biran and Naveh offer no reconstruction for the end of the line.
At the end of the line, Biran and Naveh propose the restoration of the
partially damaged 'ayin and the addition of a taw.However, they also
propose the addition of at least one divider and extra letters which they
have not specified. That is, they cannot reconstruct the end of Line 5 in
order to match up with the beginning of Line 6. If we allow for the inser-
tion of a taw and a word divider after the restored 'ayin, as Biran and
Naveh suggest, there is room enough for one letter and one word divider if
Line 3 is allowed to dictate the position of the inscription's left edge. If we
allow Line 4 to dictate the position of the inscription's left edge, we have
room enough for two letters and one word divider. However, Line 6 begins
with ayodh, the last letter of a word which is no longer extant. This being
the case, we must presume that no word divider should be added other
than the one after the taw which Biran and Naveh suggest. Thus, we have
a scenario in which either one or two letters can be added to the end of
Line 5, depending on whether we take our cue for the inscription's left
edge from Line 3 or Line 4.
accommodate at least the resh of bK"ICT with room to spare. It is odd then,
if this was the case, that the letter resh appears at the beginning of Line 4
and not at the end of Line 3. The word divider required at the end of Line
4 could be easily accommodated and there would be ample room for at
least two letters at the end of Line 5.
Thus, allowing Line 6 to dictate the left edge of the inscription permits
us to accommodate the letters needed by Biran and Naveh in previous
lines. However, the oddity about the noted shortness of Line 3 undermines
this reconstruction.
Orthographical Considerations
A major inconsistency also exists within the orthography of the text sug-
gested by Biran and Naveh. In order to maintain the connection between
the end of Line B3 and the beginning of Line A4, Biran and Naveh must
posit the lack of a word divider in the construction ^NHCTD^D ('the king of
188 The Tel Dan Inscription
Restored Letters
Biran and Naveh's reading is further undermined by the restoration of cer-
tain letters on Fragment A, which were made in the epigraphical analysis.
Line A4. Over the lacuna on Line A4, the remains of a lamed were
identified. This means that Biran and Naveh's reading of this word as ^K
('my father') cannot be upheld. Rather, the word is to be restored as
[...pn^N and regarded as fragmentary. As a result, Biran and Naveh's
reconstructed text for Line 4 cannot stand. Also, the arrangement of the
fragments is brought into question because no meaningful text can be
reconstructed with this arrangement after the restoration of the lamed over
the lacuna.
Line A6. The last extant letters on Line A6 are the remains of a mem and
lamed. In the reading proposed by Biran and Naveh, these letters must
form part of the word p^Q ('kings'). Some of the impetus for this reading
must be attributed to Yardeni, who identified the remains of a nun along
the very edge of Fragment A.13
However, closer inspection of the remains of this letter clearly shows it
to be part of a waw, not a nun. As a result, the last word of Line A6 cannot
be restored as p^Q. Thus, Biran and Naveh's reading at this point cannot
be upheld. This has implications for the restoration of the text of Line 6
between Fragment A and Fragment B as arranged by Biran and Naveh.
Their inserted text must be altered in light of the remnant of this waw at
the edge of Fragment A. Yet, no meaningful text can be proposed with this
arrangement of the fragments.
A New Arrangement
The numerous objections raised to both the text proposed by Biran and
Naveh, and their arrangement of the fragments, highlights the need for an
alternative configuration. The preceding epigraphical analysis showed that
the script of Fragment B is consistent with a scribe writing at the bottom of
a stone rather than at the top. In keeping with this observation, I propose
arranging the fragments so that Fragment B is placed below Fragment A.
A particular comment by Biran and Naveh is also pertinent to this pro-
posal of rearranging the fragments. In considering the possibilities for
arranging the fragments, they noted that Fragment B is unlikely to have
come before Fragment A in the original inscription because of the refer-
ence to the author's father in Fragment A.14 In the genre of royal memorial
inscriptions, this is indeed correct. In fact, one of the key elements which
allows us to classify the genre of the Tel Dan Inscription is the author's
reference to his father. Furthermore, the text of Fragment B does not sup-
port its placement above Fragment A for there are no references to the
author's father. Also, the content is more appropriate in the latter portions
of a royal memorial inscription than in the earlier portions prior to refer-
ences of the author's father.
Galil, however, opts for placing Fragment B above Fragment A.15 Unfor-
tunately, he does this on the surmise that Biran and Naveh are correct in
interpreting the names in the final lines of Fragment B as those of Jehoram
and Ahaziah. Galil sees the context in which these two monarchs are
named as belonging to the generation before the author of the Tel Dan
Inscription. That is, the ancient author mentioned them in a discussion of
his own father (whom he interprets as Hazael). However, it is only the
original arrangement proposed by Biran and Naveh that demands the
names be those of Jehoram and Ahaziah. As such, Galil actually bases his
new arrangement on the discredited old arrangement. Furthermore, Galil's
study lacks a detailed epigraphical examination to inform him of the line
trends of each fragment and what they entail.
These facts, coupled with the unsustainability of Biran and Naveh's
reconstructed text, leave us with only one possibility for arranging the
fragmentsnamely, that Fragment B must be placed below Fragment A.
A slight shifting of the fragments in order to match up different lines
between Fragment A and Fragment B yields no meaningful results either.
Therefore, our only option left is to place Fragment B below Fragment A.
In order to calculate just how far below Fragment A we must position
Fragment B, we must consider the dimensions of the written surface which
were calculated in the epigraphical analysis. In this section I arrived at a
minimum height of 110 cm and an approximate width of 35 cm. Line A8
was calculated at approximately 90 cm above the bottom edge of the in-
scription. This means the bottom of Fragment A was 77 cm above the
bottom edge of the inscription. We also noted that the writing on Fragment
B reflected a position towards the bottom of the stone. In writing the text
of Fragment B, the scribe had good control of his writing implement. As a
result, I suggest the scribe was not leaning forward over the stone when
writing the text of Fragment B. Rather, he was in a comfortable posture
with his arm in a fairly natural writing position. As such, I suggest placing
Fragment B within a cubit's length (about 45 cm) of the bottom edge of
the inscription. However, we have no clue as to where in the stele's width
we should place Fragment B. That is, the longitude of Fragment B is un-
known. What is certain, however, is that it cannot be placed very close to
either the left or right edge of the stele because there are no remains of any
edges.
15. G. Galil, 'A Re-Arrangement of the Fragments of the Tel Dan Inscription and
the Relations between Israel and Aram', PEQ 133 (2001), pp. 16-21.
5. Arrangement of the Fragments 191
Figure 5.4. A reconstruction of the probable position of the fragments in relation to the
whole original stele, showing Fragment A in the upper portion and Fragment B in the
lower portion (dimensions in centimetres).
Chapter 6
TEXTUAL ANALYSIS
Introductory Remarks
I turn now to an analysis of the text of the Tel Dan Inscription. Since I
propose placing Fragment B at some distance below Fragment A in the
original composition of the text, I will examine Fragment A and Fragment
B separately. The aim is to derive a meaningful text from the deciphered
characters and from those characters that can be restored or reconstructed.
In the course of this analysis, I will examine the meaning of individual
words, and address issues of grammar and syntax.
The Text
Fragment A: Transcription
[ ^]i>noh[ ] (Ai)
[ ]pD^3K^n]b[nm] (A2)
[K^D:> ^irr^Korjen (A3)
[ ^npn^'p-lKIl'Dlp^Kl (A4)
par* ]4'mp-nn-[rr>njK (A5)
p^K ]1[']pl2bnp>'Dl?Q^ (A6)
[TriKiir j-criB^a^Kvra (A?)
[^Q nn]^np>^ner^Q (A8)
[ ]-&z>Tnrr>"[ (A9)
[ ib-cn-p-iK-rr (AlO)
pMBtnrr ?]!fnpnK (All)
[HOE? ^K-OET'^'-f? (A12)
[ ]-lD2J]^rnSl3 (A13)
Fragment A: Translation
(Al) [ ] you will rule ov[er ]
(A2) [and because of the p]iou[s act] s of my father, may [?] go up [ ]
(A3) and my father will repose. May he go to [ at every]
(A4) ancient [h]earth on ground of El-Bay [tel am]
(A5) I, so Hadad would go before me [ the day-]
(A6) -s of my reign, and I would slay a kin[g] and [ thousands of cha-]
(A7) -riots and thousands of horsemen [ ]
(A8) the king of Israel, and [I] killed [him kin-]
(A9) -g of Bayt-Dawid. And [the] name of [ ]
(AlO) their land to [ ]
(All) another and to [ Jehoashr-]
(A12) -eigned over Is[rael I laid]
(A 13) siege to [Samaria ]
Fragment B: Transcription
[ Tim[. ] (Bi)
[ v^nan^htnn ] (B2)
[ -nr>]^i>bin;i[ ] (B3)
[ Tp]ft-"nn'f7!3r>?[ ] (B4)
[ ]>Q]iOZ>jBpSK[]3 ] (B5)
[ EJ]K""IDfriiJ[ ] (B6)
[ KD-D-I[ ] (B7)
[ Wvn>in'[iO* ] (B8)
194 The Tel Dan Inscription
Fragment B: Translation
(Bl) [ ]and[?]cut[ ]
(B2) [ in] his [flighting against A[? ]
(B3) [ ]?. But my king, [Hadad,] would come [ ]
(B4) [ ] Hadad made m[e] king [ ]
(B5) [ bjraver than seven [kings ]
(B6) [ ]ty captured m[en ]
(B7) [ ?[ramsonof[ ]
(B8) [ Amaz]iah son of [Joash ]
Commentary on Fragment A
Line AI
The first portion of the line is missing due to breakage of the stone. There
is room for approximately eight or nine characters before the first extant
letter. We must note the probability that this was not the first line in the
whole original inscription because of the height of the exposed surface
above this line.
[...]l>~ln[...]. Biran and Naveh misunderstood the initial markings
here as a mem.l Closer epigraphical observations reveal the markings to be
two distinct letters, a taw followed by a sin. The difficulty in interpretation
here is that we do not know if the letters "liZn form an independent lexeme
on their own. The breakage of the stone precludes us from making an
absolute judgment on this. However, we can deal with these letters as one
whole lexeme in order to determine if any meaning can be assigned to
them independently.
No noun "l!Zn is attested in Northwest Semitic languages. There is a
suggested emendation to a Hatrean text (49.3) which replaces K~ltyn with
"Itn.2 However, no meaning is offered for "icn. Alternatively, we could
see "I En as the end of a personal name with the final two letters ("IEJ) being
the noun 'prince' or 'potentate'. A less likely option is to view this as a
defective spelling for the month 'Tishri'. In order to maintain this, how-
ever, we need to posit some Mesopotamian influence over Dan and the
surrounding region so that the calendar bore Mesopotamian names. Unless
the Tel Dan Inscription was a bilingual inscription, we cannot really posit
that the author was from Mesopotamia.
1. Biran and Naveh, 'An Aramaic Stele Fragment', pp. 87, 90.
2. J. Hoftijzer and K. Jongeling, Dictionary of the North-west Semitic Inscriptions
(2 vols.; Leiden: E.J. Brill, 1995), II, p. 765.
6. Textual Analysis 195
Line A2
^[l]b[rm]. Again, the first portions of the line are missing or damaged
because of the stone's fracturing. However, traces of deliberately carved
strokes allowed for the restoration of these letters. As was seen in the
epigraphical analysis, the first fully legible character in Line A2 is a word
divider. It was possible to restore ayodh immediately before this. The pre-
ceding letters were restored on the basis of physical alignment and space,
as well as lexical and contextual sense, so that the word before the word
divider read ^n]b[rm].
The noun "TDPf is attested in the singular in later Jewish Aramaic from
Palestine with the meaning 'piety'.8 Although this would appear to be
under the influence of Biblical Hebrew, where it is far more common, an
adjectival form TOP! is known from Punic epigraphic texts.9 Contextual
analysis shows that it implies an act of kindness, often unwarranted, moti-
vated by the subject's commitment to the recipient of the action.10
<
OK'*["T]b[rD'l]. The addition of ayodh to the root "TDPI means we are
almost certainly dealing with a noun here. An imperative fem.sg. is highly
unlikely given the genre and content of the inscription. The yodh could
represent either a l.com.sg. pronominal suffix ('my piety'), or a plural
construct ('pious acts of). The following words make the first option
unlikely since subjects commonly follow verbs (see below). Therefore, we
may view 1'[~r]b[rQl] as being in construct with the following word, ""UK.
The construct chain yielded, then, is ^^''["TjbfrQl] ('and because of the
pious acts of my father').
The speaker is clearly the author of the inscription. The impetus for
adding a prefixed preposition D to the restored word "HDil comes from a
similar expression in other royal inscriptions. In Panammu II, line 19, we
find the expression "'pliDl <'3N"p"7iO ('because of my father's righteous-
ness and my righteousness'). The exact same expression appears in Bar
Rakib, lines 4 and 5. Both these stelae were produced by Bar Rakib and in
both instances, the expression precedes a statement about Bar Rakib's
ascension to the throne. Similarly, in lines 1-2 of Panammu II, we find the
expression rON-pliC ('because of his father's righteousness'). The stele
of Si'-gabbari from Nerab also contains an expression with the word
11. The syntax of verb + subject is certainly not universal, particularly when a
subject is emphasized by the use of a personal pronoun, or when a disconnected clause
is resumed by a personal pronoun. The point here is that the syntax of verb + subject is
far more common, and that there is evidence to suggest that this is upheld in the Tel
Dan Inscription.
12. Biran and Naveh, 'An Aramaic Stele Fragment', pp. 90-91.
13. The syntactical construction of subject + verb occurs occasionally in the Deir
'Alia texts. However, since they are poetic texts, we are dealing with a different genre
in which the syntax is dictated by poetic conventions. The Tel Dan Inscription is not of
poetic genre. Rather, it employs syntax more akin to prose narrative. We cannot, there-
fore, use the Deir 'Alia texts to compare with the syntax of the Tel Dan Inscription.
198 The Tel Dan Inscription
these clauses, we see the subject preceding the verb. In the stele of Sin-
zer-ibni from Nerab, we read in lines 9-10, inD^rav^DDVEmnontB
j^KQ^^Nl "[D2> ('May Sahar and Shamash and Nikkal and Nusk drag
your name and your place out of life!'). Here we see the four expressed
subjects of the verb 1PID11 placed before the verb itself. Similarly, in the
stele of Si'-gabbari from Nerab, we read in lines 9-10, "fra>^DD>"in27
lQKnnrnnN>nnnDf>:in n ('May Sahar and Nikkal and Nusk make
his dying odious, and may his posterity perish!'). Again, we see two
clauses in which the expressed subjects precede their corresponding verbs.
The syntax of the two clauses from the Mesha Stele, however, are of a
particular constructionnamely, X + qatal. This syntax, in which the
subject precedes an afformative verbal conjugation (perfect), is idiomatic
for expressing past tense with a pluperfect nuance. Since we are dealing
with a preformative verbal conjugation here in the Tel Dan Inscription, the
syntax does not really compare.
The syntax of the examples from Nerab appears quite compelling in
making us connect ""DN in Line A2 with the following verb pD\ In both
these stelae from Nerab, and in the Tel Dan Inscription, we are dealing
with preformative verbal conjugations. However, Line A5 of the Tel Dan
Inscription contains unambiguous evidence that subjects follow their corre-
sponding verbs of preformative verbal conjugations in the Tel Dan Inscrip-
tion. The preformative verbal form "[IT! possesses a prefixed waw that
separates it syntactically from the previous word. The only way in which
this verb may be read is to connect it with the following personal name,
T7!"I. We therefore have the syntactical construction in which the prefor-
mative verbal form precedes its subject. It must also be mentioned that the
verbal forms from the Nerab Stelae are jussives. Although the verb pD"* in
Line A2 may also be a jussive, the normal syntax is to place jussives before
their subjects. This is overwhelmingly demonstrated with the numerous
jussives in both Panammu I and the Sefire Treaties.14 Thus, the syntax of
the jussives in the Nerab Stelae is irregular. The beginning of Line A3 also
supports placing the verb before the subject C'HN'iacn). Thus, although
it is possible to associate "QN with pD% the possibility is remote. On con-
textual and syntactical grounds, therefore, I propose connecting "ON in
Line A2 not with the following verb, pD\ but with the preceding word,
"[7]b[rm].
14. Sefire II A.4 may be an instance in which the subject of the jussive precedes the
jussive itself. However, the text is fragmentary at this point and all other jussives
follow the conventional syntax with the subject following the verb.
6. Textual Analysis 199
All the references from other texts cited by Biran and Naveh in support
of their reading to place ''DK with pD"1 actually count against them. Two
references from the Sefire Treaties are cited (I A.5 and I C.3-4) in which
the verb pD"1 or ]pDn appears in a subordinate relative clause.15 Thus, the
syntax does not compare with that suggested by Biran and Naveh for Line
A2. Similarly, the Akkadian reference cited by them is also in a relative
clause.16 The Biblical Hebrew reference from 1 Sam. 25.13 has no ex-
pressed subject while the citation from 1 Kgs 1.40 does have an expressed
subject, but it follows the verb. Therefore, there is little syntactical support
for understanding "ON as the expressed subject of pD\
Furthermore, since Line A2 is broken off at the qoph of pD% we cannot
even be sure whether this verb is singular or plural. This adds further un-
certainty to a direct connection between "DK and pD"1. Reasonably, there-
fore, we must connect "ON with the preceding word, * [~I]b[mi], yielding a
construct expression, ^UlN^ ["T]b[rm] ('and because of the pious acts of my
father').
Two possible implications derive from this expression. Either the author
of the inscription mentioned his own piety before that of his father (in con-
trast to Bar Rakib), or he did not mention his own piety at all. If the former
case is true, then the context for the expression in Line A2 may be the
author's ascension to the throne, as it is in Panammu II (line 19) and Bar
Rakib. Alternatively, if the author of the Tel Dan Inscription did not men-
tion his own piety, we may have an expression similar to lines 1-2 of
Panammu II. There we read |Qn^n^niCO^Sn3^pliJnlQ]2^n^
nnn^> ('My father Panammu, because of his father's righteousness, did
the gods of Ya'di deliver him from destruction'). In the case of the Tel
Dan Inscription, the author would not be referring to the pious acts of his
father's father, but simply to those of his father.
References to an author's father are traditionally placed in the early
portions of an inscription. The texts of the Mesha Stele, Panammu I and
Bar Rakib all demonstrate this. The Panammu II Inscription is different in
that Panammu II was not the author. The inscription was authored by his
son, Bar Rakib, and was a memorial about his father, Panammu II. As
such, references to the author's father at the end of the Panammu II
15. Sefire I A.5 reads mEJKD jpD"1 ''T ni]H DU1 ('and with his sons who will come
up in his place'). Sefire I C.3-4 reads '"IEa ]pD' 'T HH 13^1 ""13^ ('for my son and
for my grandson who will come up in my place').
16. The Akkadian reference reads mannu sarru sa ilia arkiya ('whatever king will
arise after me'), cited from CAD, 'E', p. 123, s.v. elu.
200 The Tel Dan Inscription
Inscription must be classed differently since the author's father is the main
referent of the text. The author is only a secondary referent. We can be
sure that we do not have a similar case in the Tel Dan Inscription because
the author refers to his killing of others in Line A6. This would be appro-
priate if the author was indeed the main referent of the text. It is highly
unlikely that the author would mention his own personal exploits in isola-
tion or in clear distinction to those of his father if his father were the main
referent of the text.17
[.. .]pD\ This verb confirms that the language of the inscription is Ara-
maic since the root of this verb, Vp^D, is attested only in Aramaic. Further-
more, the verbal form allows us to classify the language more specifically
as Old Aramaic. Later strands of Aramaic retain the lamed of Vp^D in the
imperfect of the Pe 'al conjugation, and only in the Haph 'el or Aph 'el con-
jugations does the lamed elide.18 The lack of a heh or 'aleph preformative
in the current form would seem to preclude the verb from being in one of
these conjugations. The dialect of the Deir 'Alia texts (if it can even be
classed as Aramaic), and that of the Sam'alian texts from Zenjirli, are the
only known dialects in which the preformative heh of the Haph 'el conjuga-
tion is lost in the imperfect.19 The close similarity between the dialect of
the Tel Dan Inscription and the Sefire Treaties suggests that it is likely the
heh preformative of the Haph 'el conjugation was retained in the Tel Dan
Inscription, as it is in the Sefire Treaties. There is no prefixed mem to sug-
gest an infinitive form.
This leaves the Pe 'al or Pa 'el conjugations as options for this verb.
Elsewhere, the verb is only attested once in the Pa 'el stem as a perfect in
Egyptian Aramaic texts, but the meaning is ambiguous. It could mean 'to
17. Kilamuwa mentions the ineffectiveness of his royal predecessors in his own
stele. However, none of these predecessors are the main referent of Kilamuwa's stele.
Rather, Kilamuwa himself is the main referent.
18. See Hoftijzer and Jongeling, Dictionary of the North-west Semitic Inscriptions,
II, pp. 788-90.
19. See J. Hoftijzer and G. van der Kooij (eds.), Aramaic Texts from Deir 'Alia (Lei-
den: E.J. Brill, 1976), p. 293; Gibson, Textbook of Syrian Semitic Inscriptions, II, p. 63.
There is also a suggestion of an early Aph 'el form in Sefire III.3see Fitzmyer, The
Aramaic Inscriptions of Sefire, p. 145. However, others have conjectured that this
isolated Aph 'el form is a scribal error. See S. Segert, 'Zur Schrift und Orthographic der
altaramaischen Stelen von Sure', ArOr 32 (1964), pp. 110-26 (121); J.C. Greenfield,
'Studies in West Semitic Inscriptions. I. Stylistic Aspects of the Sefire Treaty Inscrip-
tions', AcOr 29 (1965), pp. 1-18; Degen, Altaramaische Grammatik, p. 19 n. 79.
6. Textual Analysis 201
24. J.A. Hackett, The Balaam Text from Deir 'Alia (Chico, CA: Scholars Press,
1980), pp. 123-24.
25. Contra Hoftijzer in Hoftijzer and van der Kooij (eds.), Aramaic Texts From
Deir 'Alia, pp. 296-98.
26. Cf. J.A. Emerton, 'New Evidence for the Use of Waw Consecutive in Aramaic',
VT44 (1994), pp. 255-58; V. Sasson, 'Some Observations on the Use and Original
Purpose of the Waw Consecutive in Old Aramaic and Biblical Hebrew', VT41 (1997),
pp. 111-27; V. DeCaen, 'The Morphosyntactic Argument for the waw Consecutive in
Old Aramaic', VT51 (2001), pp. 381-85.
27. This will be discussed below.
28. Muraoka's work appears in four articles. The first, 'Linguistic Notes on the
Aramaic Inscription from Tel Dan', IEJ 45 (1995), pp. 19-21, accompanied the
publication of the find of Fragments B1 and B2 by Biran and Naveh. The second, 'The
Tel Dan Inscription and Aramaic/Hebrew Tenses', Abr-Nahrain 33(1995), pp.113-15,
was written in response to Tropper's appraisal of the Tel Dan fragments (see Tropper,
'Palaographische und linguistische Anmerkungen'). Muraoka's third article, 'Again on
the Tel Dan Inscription and the Northwest Semitic Verb Tenses', ZAH11 (1998), pp.
74-81, was also written in reponse to an article by Tropper entitled, 'Aramaisches
wyqtlund hebraisches wayyiqtoF, UF2& (1997), pp. 633-45. Along with M. Rogland,
Muraoka wrote a fourth article, 'The waw Consecutive in Old Aramaic? A Rejoinder
to Victor Sasson', FT48 (1998), pp. 99-104, and then a fifth, 'The Prefix Conjugation
in Circumstantial Clauses in the Tel Dan Inscription?', FT51 (2001), pp. 389-92.
29. Muraoka, 'Linguistic Notes', p. 20. Biran and Naveh, however, nowhere argue
for the use of a copulative waw rather than a waw consecutive.
6. Textual Analysis 203
33. Gibson cites examples of preformative verbal conjugations with past tense mean-
ing in Biblical Aramaic; see Gibson, Textbook of Syrian Semitic Inscriptions, II, p. 15.
However, only one preformative verbal conjugation could be considered a preterite
(Din1, Dan. 4.31, 33). However, as the LXX demonstrates, there are numerous textual
difficulties with these verses. Nevertheless, the use of preformative verbal conjugations
in these two verses is not indicative of past tense, but an action simultaneous with that of
another. Thus, they do not carry the concept of consecutive action, but of simultaneous
action. See also F. Rosenthal, A Grammar of Biblical Aramaic (Wiesbaden: Otto Harras-
sowitz, 1963), 178.
34. See Hackett, The Balaam Text from Deir 'Alia, pp. 123-24.
6. Textual Analysis 205
Despite the fact that "OK and pD11 are in separate clauses, it is the context
which is under scrutiny. In this regard, Line A3 must also be considered.
read this as a reference to the author's father having died and gone 'to his
house of eternity'.52 However, "]T should not be interpreted as a preterite,
but as a jussive in an optative expression.
52. nQbUHTIl*1^; see Biran and Naveh, 'An Aramaic Stele Fragment', pp. 91-92.
53. Biran and Naveh, 'An Aramaic Stele Fragment', p. 92.
54. Biran and Naveh, 'The Tel Dan Inscription', pp. 14-15.
55. Garbini, 'L'iscrizione aramaica di Tel Dan', p. 463. See also Hoftijzer and
Jongeling, Dictionary of the North-west Semitic Inscriptions, II, pp. 986-92.
56. Donner and Rollig, Kanaandische und Aramdische Inschriften, 43.12.
57. Compare the use of Dip in Biblical Hebrew, Pss. 74.2; 119.152.
6. Textual Analysis 209
58. E.A. Knauf, A. De Pury and T.R. Romer,' *BaytDawTdou *BaytDod?\ BN12
(1994), pp. 60-69(62-63,68).
59. Hoftijzer and Jongeling, Dictionary of the North-west Semitic Inscriptions, II,
pp. 986-87.
60. Hoftijzer and Jongeling, Dictionary of the North-west Semitic Inscriptions, II,
p. 987.
61. Cryer, 'On the Recently-Discovered "House of David" Inscription', pp. 16,18.
62. The word is also used as a symbolic name for Jerusalem in Isa. 29.1, 2, 7.
63. This suggestion is based on Mesha's boast of having brought the rnn^KlN
('the hearth of its dwd'} from 'Ataroth to the shrine of Kemosh in Qiryat.
210 The Tel Dan Inscription
context very well. It also compares with Panammu I, lines 15-18 and 21 -22,
in which Panammu I entreats his successors to bless his memory and con-
dition in the afterlife when they sacrifice to Hadad. That is, the memory of
the deceased ancestor was to be blessed in cultic rites. The restoration of
['TOD] before D"7p>L?K"l[N] is far from certain. However, it would not be
inappropriate in the context.
[...^Nn^D't'N'pnND. Biran and Naveh understood p~)K as referring to
the land of the author's father.64 However, since we restored a lamedover
the lacuna in Line A4, this reading cannot stand. I propose connecting these
two words to the preceding clause. We must consider [.. .pn't'N as fragmen-
tary since vD^N is unattested as a genuine root in Northwest Semitic lan-
guages. Although Vl^K is attested as an orthographical variant of V*]1^
('to instruct, incite'),65 we cannot make good contextual sense of this root
here. In addition, we see the noun "'S^N in Line A7, making the interchange
ofpeh with beth highly improbable in the Tel Dan Inscription. If the clause
ended with p"IN3, we might expect to find this noun in the emphatic state
with the post-positional definite article (Kp"lKH). It may be argued that
since the definite article is unattested in the Tel Dan Inscription that we
cannot know with any certainty whether it was employed at all,66 or
whether the definite article was not marked with a prefixed definite article
as in Hebrew. It has been argued that Line B4 preserves a prefixed definite
article in the lexeme "f^QI"!.67 This, however, is highly unlikely since the
vocabulary clearly marks the language as Old Aramaic. Furthermore, the
script shows distinct Syrian influence, which would point to the definite
article being the post-positional 'aleph. Although this particular point is
not an all-encompassing argument, it certainly lends weight to the notion
that we are dealing here with a conventional dialect of Old Aramaic, simi-
lar to that in the Sefire Treaties. Therefore, the lack of a post-positional
definite article here most probably indicates that p"lKD is in construct with
the following word.
With this understanding, I have restored [.. .J^R to read ^WTn^N ('El-
Baytel'). No other meaningful alternative for understanding this lexeme
can be found. 'El-Baytel' is the hypostasis of the deity El, who came later
64. Biran and Naveh, 'An Aramaic Stele Fragment', pp. 90, 92; idem, 'The Tel
Dan Inscription', pp. 13-15.
65. Panammu I, line 34. See also Hoftijzer and Jongeling, Dictionary of the North-
west Semitic Inscriptions, I, p. 64.
66. Compare the lack of a definite article in Sam'alian Aramaic.
67. Cryer, 'King Hadad', p. 232.
6. Textual Analysis 211
Line A5
[...]^Qlp~nn"[lT>n]N. Biran and Naveh understood the first word in
Line A5 as the l.com.sg. independent personal pronoun, T. Originally,
before the discovery of Fragment B, the context was understood as a state-
ment concerning the author's humility, as in Zakkur I, line 2, n38n]I>^N
('I am a humble man').77 After the discovery of Fragment B, this was
altered to coincide with the arrangement of the fragments and the text pro-
posed by Biran and Naveh. The new context was understood to be the
implies certain victory over the enemy. We may infer from this that the
author of the Tel Dan Stele considered himself favoured by Hadad to
whom he ascribed or dedicated his military exploits. The use of the habit-
ual imperfect ~[i~H implies that the author was a regular military cam-
paigner who, in his own estimation, enjoyed a good deal of success.
80. Biran and Naveh, 'An Aramaic Stele Fragment from Tel Dan', p. 92.
81. So Fitzmyer, The Aramaic Inscriptions of Sefire, p. 52. However, only the
initial mem is preserved. Gibson disagrees and reconstructs this word as [TO^p. See
Gibson, Textbook of Syrian Semitic Inscriptions, II, pp. 32, 43.
82. See further Fitzmyer, The Aramaic Inscriptions of Sefire, pp. 117-18.
214 The Tel Dan Inscription
83. See further, W.R. Garr, Dialect Geography of Syria-Palestine, 1000-586 BCE
(Philadelphia: University of Pennsylvania Press, 1985), pp. 44-45, 72 n. 168.
84. See Sefire I B.27; II B.8-9; III.l 1, 18, 21.
85. See Panammu II, line 8.
86. Biran and Naveh, 'The Tel Dan Inscription', p. 16.
87. Compare the use of fTN in Lines A9-10 and Line B4.
6. Textual Analysis 215
Since ["]]^b is followed by a word prefixed with waw, I expect that the
author went on to explain exactly how he dealt with a defeated king, or
otherwise how he killed other persons, such as commanders, soldiers, or
even entire armies. Presumably, this explanation made reference to the
defeat and plunder of the defeated king's army. Line A7 supports this
understanding. The author thus portrays himself as a fearsome warrior
before whom no one could stand. Not even kings could endure encounters
with him for he was a slayer of kings and an annihilator of armies.
91. The Samaria Ostraca, dated to just after the Tel Dan Inscription, prove that
Hebrew was the language of Israel at this time.
92. The reasoning behind this restoration will be discussed in the following chapter,
since it is based primarily on historical considerations rather than textual or epigraphic
reasons.
6. Textual Analysis 217
should, therefore, be applied here. We must translate this verb, then, with
past-tense meaning as 'and [I] killed [him]\
Line A9
TnrrD. Without doubt, Line A9 has caused the greatest controversy aris-
ing from the Tel Dan Inscription. From Biran and Naveh's initial publica-
tion of Fragment A, debate has raged over the interpretation of the lexeme
Tnrf 3. Biran and Naveh first posited the theory that it should be inter-
preted as 'House of David', referring to 'the dynastic name of the kingdom
of Judah'.93 A parallel was drawn to the Assyrian designations of various
small states in the Levant, such as BitHumri ('House of Omri' = Israel),
BitAgusi ('House of Agusi' = Arpad) and Bit Haza 'Hi ('House of HazaeP
= Aram-Damascus). These designations were styled after the names of
prominent rulers within these states, rather than the actual name of the
state. That is, certain states came to be known after the name of a major
dynasty. This line of interpretation was followed numerous scholars, such
as Ahituv,94 Kallai,95 Andersen,96 Kitchen,97 Lemaire,98 Noll,99 Puech,100
Rainey,101 Rendsburg,102 Schniedewind,103 Tropper,104 Wesselius,105 and
93. Biran and Naveh, 'An Aramaic Stele Fragment', pp. 93, 95-96.
94. S. Ahituv, 'Suzerain or Vassal? Notes on the Aramaic Inscription from Tel
Dan', IEJ43 (1993), pp. 246-47.
95. Z. Kallai, 'The King of Israel and the House of David', 7E/43 (1993), p. 248.
96. F.I. Andersen, 'I Have Called You By Name...', BH 34 (1998), pp. 37-52
(44-45).
97. K.A. Kitchen, 'A Possible Mention of David in the Late Tenth Century BCE,
and Deity *Dod as Dead as the Dodo?', JSOT16 (1997), pp. 29-44 (38-39).
98. A. Lemaire,' "House of David" Restored'; idem, 'The Tel Dan Stela as a Piece
of Royal Historiography', JSOT81 (1998), pp. 3-14.
99. Noll, 'The God Who is Among the Danites'.
100. Puech, 'La Stele Arameene de Dan'.
101. A. Rainey, 'The "House of David" and the House of the Deconstructionists',
BARev 20.6(1994), p. 41.
102. G.A. Rendsburg, 'On the Writing TnfTD in the Aramaic Inscription from Tel
Dan', IEJ45 (1995), pp. 22-25.
103. Schniedewind, 'Tel Dan Stela'.
104. Tropper, 'Eine altaramaische Steleninschrift aus Dan'; idem, 'Palaographische
und linguistische Anmerkungen'.
105. J.W. Wesselius, 'De eerste koningsinscriptie uit het oude Israel: Een nieuwe
visie op de Tel Dan-inscriptie', NTT53 (1999), pp. 177-90,243, translated into English
as 'The First Royal Inscription from Ancient Israel: The Tel Dan Inscription Recon-
sidered', SJOT 13 (1999), pp. 163-86. A response was offered by B. Becking, 'Did
Jehu Write the Tel Dan Inscription?', SJOT 13 (1999), pp. 187-201.
218 The Tel Dan Inscription
Yamada.106 On the other hand, there were those scholars who insisted that
TnJTn could not be interpreted as 'House of David', such as Cryer,107
Davies,108 Knauf, De Pury and Romer,109 Lehmann and Reichel,110
Lemche111 and Thompson.112 It is obvious that this conundrum has yet to
be resolved.
The crux for interpreting the lexeme TITTD lies in the fact that there is
no word divider between the seeming two parts, JT3 and 111. This sug-
gests that the lexeme incorporates only one idea rather than two separate
ideas, and is to be understood as a single concept or entity. This is con-
firmed by the fact that elsewhere in the Tel Dan Inscription, construct
expressions are used to denote two or more concepts that are both indi-
vidually exclusive, yet connected genitivally in the given context. As a
result, a word divider is used to demarcate the separate parts of a construct
expression. For example, in Line A4 we see the construct expression, p"IN3
[^Nflpn^K* ('on ground of El-Baytel'). Similarly, in Line A8 we see the
expression, ^KIET "]L?Q ('the king of Israel'). In Line A7, we also have the
expression, cnB^BvNl ('and thousands of horsemen').113 We even see a
surplus use of word dividers in Line Al 0 where a noun is demarcated from
a pronominal suffix in the expression D!>p"IK ('their land').114 Thus, there
can be no doubt that the lexeme TnjT3 denotes one idea rather than two
120. H.M. Barstad and B. Becking, 'Does the Stele from Tel-Dan Refer to a Deity
Dod?', 57V77 (1995), pp. 5-12.
121. Certain names in Ancient Arabic inscriptions from c. 500 BCE contain the the-
ophoric elements 'Dad' and 'Dadat'. These are most likely to be refer epithets for other
deities rather than deities in their own right. See Hoftijzer and Jongeling, Dictionary of
the North-west Semitic Inscriptions, I, p. 493.
122. See Chapter 7 for further discussion of this evolution.
123. Knauf, De Pury and Romer, in distinction to Lehman and Reichel, do not
propose understanding the lexeme D27N1 as a reference to the deity Ashim(a). Rather,
Knauf, De Pury and Romer interpret this word as a consecutive imperfect meaning,
'and I setup'.
6. Textual Analysis 221
of the verb. That is, the syntax would require 'BaytDod' to be the item
poured out as a libation, rather than the indirect object to whom the
libation was made, or the item on or inside which the libation was made.
The only way this interpretation can be avoided is to supply a preposition
immediately before "inn"13. The prefixed preposition ^ is needed to under-
stand 'BaytDod' as the deity to whom libation was made. Alternatively,
the prefixed preposition H is required to understand 'BaytDod' as a temple
or shrine at which the libation was made. Since neither of these preposi-
tions is present in the text, it is only with a great stretch of normal gram-
matical principles that we can avoid considering TnJT3 as the direct
object rather than the indirect object of a preceding verb. Even still, if we
interpret TOfTD as a cultic object of some sort, as Lehmann and Reichel
do, the waw conjunction prefixed to DCft* ('Ashim') further obfuscates the
syntax so that deriving a meaningful sense from the text is elusive. It is
abundantly clear, then, that the suggestions of Lehmann and Reichel, as
well as those of Knauf, De Pury and Romer, have too many holes in them
to be considered sustainable theories in regards to the word TITrfD.
The suggestion that TnJTD be translated as 'House of David' is also
fraught with difficulties. The fact that the words JTD and 111 have been
combined into one composite lexeme means that the term was most prob-
ably not understood as a construct expression, 'House of David'. That is,
the author was certainly not referring to a Davidic dynasty that ruled the
small ancient state of Judah. The orthography of the text indicates that such
an expression would most certainly have been rendered by two separate
words, TnJT3. As such, the translation 'House of David' is impossible.
Almost as a concession to this theory, however, is the fact that TnJTD
might refer to an actual state entity. We may consider TITTD as a singular
proper noun to be rendered 'Bayt-Dawid'. In this regard, the author would
be referring to the name of a small state. Although such a state would have
received its name from a prominent dynasty that ruled it, the label inn1'3
would not refer to the dynasty itself, but to the actual country ruled by that
dynasty. That is, Tnff H could be interpreted as a political term for a state
entity. This is how such state labels of the type '5ff-PN' are used in Assyr-
ian texts. Such usage is to be clearly distinguished from the proposal that
TnrTD be translated as a dynastic name, 'House of David'. Nevertheless,
this hypothesis by default incorporates an understanding that a 'House of
David' at one time ruled a state entity.
However, the lexeme Tnmi may just as reasonably be understood to
refer to a geographical entity, rather than a political entity. That is, TniTH
222 The Tel Dan Inscription
would present a parallel with the lexeme Tnmi in the Tel Dan Inscrip-
tion, we again encounter the problem of word division, for the Palmyrene
expression is a construct expression composed of two distinct words. We
might also wonder what a reference to a 'cookhouse' in the Tel Dan In-
scription would entail for both the context and general understanding of
the inscription.
Inevitably, we must interpret "PUT 13 in a manner that is meaningful to
the immediate context within the inscription. Thus, meaningful translation
and interpretation of the lines surrounding Line A9 is the best approach to
understanding the term TnrTQ in the Tel Dan Inscription. Lines A7-8, as
discussed above, refer to the author's encounter with the king of Israel,
from whom he appears to have struck down or captured vast numbers of
military units. I presume that the author killed the king of Israel (Line A8).
In Line A10, after the reference to TnrTQ, the author refers to D!>p~lK
('their land'). Presumably, one of the entities entailed in the S.masc.pl.
pronoun DH is the king of Israel. However, since the pronoun is plural, the
author obviously envisages more than just the king of Israel here.
This gives us two options for understanding this plural pronoun. First,
the author may have had in mind both the king of Israel and his armies,
which the author boasts of having defeated or captured. Thus, 'their land'
may refer to the land of the king of Israel and his armies. It is question-
able, however, whether a king's armies would have been associated with
him in the expression, 'their land', particularly since armies were tradi-
tionally seen as belonging to the king and, therefore, on a par with the land
rather than with the king.128 This leads us to the second, more likely
optionnamely, that the author here envisaged at least one other notable
personage with comparable status to the king of Israel. That is, at least one
other monarch was associated with the king of Israel as the author's
enemy, and hence the author's usage of the S.masc.pl. independent per-
sonal pronoun, DH in Line A10.
Given that the reference to "TlinTl appears in the line before this pro-
noun, we may understand THfTD as the name of a state or city whose
leader was associated with the king of Israel. The content of the Tel Dan
Inscription, corroborated with other sources, leads us to conclude that the
name TnjTD was the Aramaic equivalent of Hebrew "111 "I1117 ('City of
David' or "Ir-Dawid'). That is, "inrvn is best understood as a reference
128. E.g. Shalmaneser Ill's account of the Battle of Qarqar (c. 853 BCE) in which the
enemy units listed as enemies belong to the respective enemy kings. See ANET, pp.
278-80.
224 The Tel Dan Inscription
14.5. There are, however, numerous difficulties associated with this pas-
sage. It is far better to compare the separation of the pronominal suffix DH
here in Line A10 with the same phenomenon in other Northwest Semitic
inscriptions. Line 18 of the Mesha Stele has the expression DnoriDNl
('and I dragged them'), Zakkur A.9 contains the expression Dnm3nQl
('and their armies'), while Bar Rakib B.7 has the expression, D!"[ncn2
('their souls'). Thus, this is not an isolated instance of separating a long
pronominal suffix from its noun. From the known occurrences, it may be
that this feature is reserved for feminine nouns with plural pronominal
suffixes.
The scope of DPI here should include the king of Israel, and presumably
also the king of Bayt-Dawid. If the king of Bayt-Dawid is indeed included
in the scope of DI~I, it is interesting that the author uses only a singular
noun, plK ('land') in mentioning their territory. This may present a case
for understanding the scope of Dil as including only the king of Israel and
his military units. More probably, the author simply regarded the territo-
ries belonging to Israel and Bayt-Dawid as a single geographic unit and
referred to them here as a single terrain rather than in the context of sepa-
rate states or countries. This suggests that Israel and Bayt-Dawid were
neighbours in the highlands of Palestine. The reference to a single terrain
is all the more understandable since there are no outstanding natural bor-
ders to demarcate different states within the highlands of Palestine south
of the Jezreel Valley. The terrain is fluidly composed of hills and valleys
from the edge of the Jezreel southward to the Negev. We may parallel
such geographic consociation with the Assyrian term for all the states west
of the Euphrates, Amurru.150
The preservation of a lamed at the extant end of Line A10 leads me to
believe the author was here claiming to have laid waste the land belonging
to the king of Israel and the king of Bayt-Dawid by turning their land 'to'
ruin. Biran and Naveh suggested that the word after the lamed here may
have been ]Q5T ('desolation').151 This word is employed in Sefire I A.32
in a curse formula. The suggestion is certainly plausible, but is still only
speculation.
Line All
jinK. This adjective, 'another', is most probably used attributively of a
preceding singular masculine noun written at the end of Line A10, but
over Is[rael]', but this is less natural than reading '[r]eigned over Is[rael]',
which is to be preferred.
Unfortunately, the subject of the verb "[^[Q] is no longer extant, though
it was presumably named at the end of Line A l l . What is clear, however,
is that if "f?[ft] is a verbal form, then the author cannot be the subject, for
we would expect a taw to have been suffixed to the verb to indicate the
first person. The content of the inscription also precludes the author from
being the king of Israel.153 Thus, we should regard ~j^[Q] as a verb in the
third person, which forms part of a statement mentioning a new king of
Israel who presumably succeeded the king whom the author had killed.
The name of this new Israelite king has been reconstructed as [2)81 IT] on
historical grounds.154
A further point of interest is that the subject of the verb, "f?[!2], actually
precedes the verb. This may have been done to emphasize the name of the
new Israelite king. Alternatively, the syntax might indicate that ~[^[Q] is to
be regarded as a participle. The same word order appears in the Mesha
Stele, lines 4-5, with the phrase ^^~l^~fiT'D''"lD^. However, there also,
there is some ambiguity as to whether "pE is a verbal form or a noun.
153. Contra J.W. Wesselius, who regards Jehu, king of Israel (841-814 BCE), as the
author of the inscription. There are, however, innumerable problems with this theory,
such as the choice of Aramaic for the inscription rather than Hebrew, the seemingly
wavering perspectives Jehu had of himself within the one inscription, the fact that the
theory depends on the erroneous arrangement of the fragments, and the lack of certain
epigraphical and archaeological considerations. See Wesselius, 'De eerste konings-
inscriptie uit het oude Israel' (or the English translation, 'The First Royal Inscription
from Ancient Israel').
154. Discussion of these historical considerations is reserved for the following
chapter.
155. Compare Zakkur A.9: "pm^JJ"lUI3-L?^3l7DL?D1iat01 ('and all these kings
laid siege to Hadrakh').
156. Compare Zakkur A. 15:1iIQ^llL'i:-lnQ'1TL'^''3^QL'D ('all these kings who
have forced a siege against you').
230 The Tel Dan Inscription
Commentary on Fragment B
Line Bl
[.. .pm [...]. Although no word divider is preserved before this word, the
waw leads me to believe that there originally was a word divider before it.
Thus, the waw is to be treated as a conjunction. All commentators have
understood this word as a verb derived from V"1T3 ('to cut'). However, no
commentator has considered the possibility that this may be a reference to
the city of Gezer, located in the Shephelah, perhaps mentioned in the con-
text of a list of towns conquered by the author. There is, however, no way
of checking this interpretation. This is further highlighted by the fact that
the line breaks off at the letter resh, leaving us ignorant as to whether or
not there were any extra letters in this word. If there originally were extra
letters, we should then interpret this word as a verb rather than a proper
noun. Unfortunately, we have no way of knowing if there were extra let-
ters without extra fragments of the stele.
This interpretational dilemma should be approached by appeal to both
context and syntactical convention. Line B2 is a reference to a battle with
the author probably being the object. That is, the author was not the aggres-
sor in this particular battle, so it is most likely that the author was the
defender here. This being the case, this particular battle probably occurred
in the author's own territory rather than in an opponent's territory. Since
the author composed in Aramaic, we should probably not regard Gezer as
the location of this battle, for he would have to be defending Gezer, in
which case we would expect the author to have written in a Canaanite
dialect, probably Hebrew. One wonders hypothetically how an account of
Line B2
[.. .?]KD"i"[QnLT![rQ...]. The first extant word, despite being in fragmen-
tary condition, is understood as a verb form from Von^ ('to fight/battle'),
also known from Moabite and Hebrew. The remnant of a taw at the
beginning of the line is the tow-infix of the Ithpe 'el conjugation, hence the
restoration of heh before it. The taw-infix is also present in the Moabite
forms from the Mesha Stele.160 However, the Moabite form employs a
metathesis between the tow-infix and the first radical, lamed. This does not
occur here in the Aramaic of the Tel Dan Inscription. Rather, the -Tin
prefix is kept as a cohesive element of the conjugation. The restoration of
the prefixed preposition D is prompted by the presence of a 3.masc.sg.
pronominal suffix, H- ('his'). We are, therefore, dealing with an Ithpe 'el
infinitive.161 A similar word, HQnn^rn, which is identical except for the
metathesis, is found on the Mesha Stele, line 19. The reference to a third
person here ('his fighting') is most likely to one of the author's enemies,
probably intended as a defamatory statement of the enemy's unjustified
aggression.
The prefixed preposition D on the following word should probably be
understood in the sense of 'against'.162 However, this is not absolutely
certain. This preposition is idiomatically employed with verbs from A/DPI1?
in the Mesha Stele to indicate the object fought against. Biran andNaveh,
however, opt for translating this preposition as 'at', indicating the location
159. See Hoftijzer and Jongeling, Dictionary of the North-west Semitic Inscriptions,
I, p. 220.
160. See Mesha Stele, lines 11, 15, 19, 32.
161. See the discussion in the Grammatical Survey later in this chapter.
162. Compare Becking, 'The Second Danite Inscription', p. 25.
232 The Tel Dan Inscription
of battle.163 Just which sense of the preposition the author intended here is
slightly ambiguous because the word to which the preposition is prefixed
is fragmentary. Thus, we do not know whether it indicated a location or a
person. Certainly, if it indicated a person, we must translate the preposi-
tion as 'against'. On the other hand, if it indicated a location, we could
translate it as either 'against' or 'at'. The evidence of the Mesha Stele sug-
gests we translate it 'against'. However, in the Mesha Stele we have a
wider definite context allowing us to make this decision. The preposition
3 is, however, used universally in this sense when preceded by a form
of Van1?. Therefore, we may suggest translating the preposition here as
'against', but we cannot make this a definite decision.
In the earlier epigraphical analysis of Fragment B, it was observed that
four letters stood as candidates for restoration of the letter following -83.
These were beth, yodh, lamed and sin. It is not possible to be any more
precise than this because of the fragmentary nature of the stone. Therefore,
we should not hazard a guess as to what this word could literally have
been, especially as it would colour our understanding of the historical
circumstances behind the inscription. Such indefinite reconstruction must
be avoided for sound historical appraisal.
Line B3
[T"[n]''DL?QL?in. The first complete word is a preformative verbal
conjugation prefixed with a waw conjunction. Biran and Naveh correctly
took this verb as a derivation from V^IJ ('to come, enter').164 The objec-
tions to this, which Becking raises,165 have more to do with Biran and
Naveh's arrangement of the fragments and the difficulty in connecting Line
B3 with the beginning of Line A4 than with translation of this verb. Beck-
ing's objections in regard to Biran and Naveh's arrangement and reading
are appropriate, but the new text created by rearranging the fragments does
not necessitate his interpretation that ^ITl is derived from Vil^^ ('to go
up'). The root, Vilbu, is common in Phoenician, Punic and Hebrew, but
only enters Aramaic at a much later stage in the Targums, under the influ-
ence of Hebrew. Also, the concept of 'going up' is conveyed by Vp^D,
which has already been seen in the Tel Dan Inscription in Line A2. Fur-
thermore, seeing the lexeme ^ITl as stemming from Vil^^ creates the false
Here in Line B3, the author may be claiming that Hadad would always
'come' to his aid. This gloss is particularly suitable in the light of Line B2,
which mentions an enemy's hostility. I suggest that the waw conjunction
prefixed to blT be understood with adversative force, 'but'.
Line B4
p!T]&"nn~[b>ftn. We have understood the first extant word of Line B4 as
a Haph 'el perfect of V"f^Q. This interpretation is prompted by two factors.
First, it was observed in the epigraphical analysis of Fragment B that the
two markings along the right edge at the beginning of Line B4 were likely
to be the remnants of either the letter gimel, heh or samekh, followed by a
word divider. Second, the context suggests the use of a perfect verb rather
than an imperfect. The phrase thus reads, 'Hadad made m[e] king', indi-
cating a singular action in past tense. This compares with the speech of
Ba'alshamayn in Zakkur A. 13, [~[rB]^Qnn]K ('it was I who made you
king'). Here in the Tel Dan Inscription, though, the king is the speaker,
rather than the deity.
Cryer's suggestion that the heh of "[^EH represents the definite article of
Canaanite dialects169 should be dismissed, particularly since the rest of the
Tel Dan Inscription reveals traits that clearly belong to conventional Old
Aramaic. Cryer's motivation for his suggestion stems from his perception
that Fragment A and Fragment B belong to two separate inscriptions and
from his dismissal of the historical interpretation offered by Biran and
Naveh.170
The lack of a waw conjunction prefixed to ~f^Qn probably indicates that
this lexeme occurs in mid-sentence rather than at the beginning of a fresh
sentence. Exactly what preceded the word "[^QH is uncertain. Our epi-
graphical analysis demonstrated that the last letter of the preceding word
could only have been a gimel, heh orsamekh. If we understand this letter as
a heh, it could be a 3.masc.sg. pronominal suffix. This could refer to either
Hadad or perhaps one of the author's enemies. However, the markings are
too ambiguous to permit confident reconstruction of the sentence. What
we may say, however, is that the words preceding ~[^QH were likely to be
the reason given by the author for why Hadad made him king.
The use of the object marker to carry the object pronoun, as opposed to
an object pronoun suffixed to the verb, may indicate an emphasis on the
object.171 However, since the object marker does not precede the verb, the
emphasis is not as strong as it could have been.172
Line B5
[ . . .]Q>jQpSK[...]. Due to the fragmentary context, it is extremely diffi-
cult to interpret this line. First, we do not know whether pSK is a complete
lexeme or whether other letters preceded it. Second, there are four differ-
ent interpretations that may be offered, of which only the first two have
previously been entertained.
1. The first reading triggered by the fragments is to consider pSN as
the Pe'al imperfect l.com.sg. of VpSDK ('to go forth'), as Biran
and Naveh have done.173 This reading is prompted by the lines of
the fragment's breakage, as well as the following lexeme, JQ
('from'). Pursuing this line of interpretation, the context does not
permit us to verify whether the form is an imperfect with future
tense ('I will go forth'), a habitual imperfect ('I would go forth'),
or even a cohortative ('May I go forth'). It is certainly not an
example of a consecutive imperfect as Biran and Naveh pro-
posed. Yet, since the previous line is certainly seen in past tense,
the verb here would most likely be a habitual imperfect, with the
text reading, 'I would go forth from seven[...]'. However, since
Fragment B belongs to the lower portions of the stele where
blessings and curses most often appear, we cannot dismiss a
future or cohortative sense. The oddity with this reading lies in
the interpretation of the word [.. .]ID& in the context. Biran and
Naveh's tentative suggestions, 'And I went forth (to war) outside
the seven (?) [districts] of my kingdom',174 seem rather awkward
and cumbersome, in spite the use of a consecutive imperfect here.
It is also motivated by their proposed arrangement of the frag-
ments. Nevertheless, we must keep the thrust of this suggestion
because of uncertainty involved in the highly damaged context.
Line B6
[.. .]N''"O*j&[...]. Biran and Naveh reconstructed the first extant letters
on Line B6 as the end of the word fJ&pCJ] ('70').180 However, equally
plausible are JS[T1K] ('40') and j [ED] ('90'), perhaps as the last part of
a larger number. The difficulty with the following word, "HDK, is whether
to interpret it as an active or a passive participle. Biran and Naveh opted
for the active, compelled by their arrangement of the fragments to read the
cumbersome expression, '70 kings who harnessed thousands of chariots
and thousands of horsemen'.!81 Thus, their arrangement also inspired their
interpretation of the single preserved 'aleph at the end of Line B6 as
pS^jft ('thousands of). Yet, since Biran and Naveh's arrangement is
dubious, we are not limited to this interpretation, plausible though it is in
the new arrangement. If we take "HDN as a passive, then we should trans-
late the lexeme as 'captured' or 'prisoners of. This allows us to interpret
the single 'aleph at the end of the line as perhaps the first letter of the
plural noun, []C?]]K ('men'), or the collective noun, [ti?]K ('men'), or even
as the first letter of a toponym, such as [D~1]N ('Aram'), or [D"T]N ('Edom').
As a tentative interpretation, I suggest 'captured men', but with the realiza-
tion that the context is too damaged to provide a definite interpretation.
LineB7
[.. .]"QD~1[...]. The final two lines of Fragment B have sparked nearly as
much controversy as Line A9. Biran and Naveh unequivocally recon-
structed this line as [nKn]--Q-D1['irr] ('Jehoram, son of Ahab').182 The
majority of objections raised to this interpretation were in regards to the
historicizing that it entails.183 Indeed, it must be said that Biran and
Naveh's interpretation represents a fairly subjective approach. Despite the
tantalizing prospect of identifying a known historical figure here, we must
180. Biran and Naveh, 'The Tel Dan Inscription', pp. 13, 16.
181. Biran and Naveh, 'The Tel Dan Inscription', p. 16.
182. Biran and Naveh, 'The Tel Dan Inscription', pp. 13, 16-17.
183. E.g. Cryer, 'King Hadad', p. 234.
238 The Tel Dan Inscription
188. E.g. ~nmn on the Melqart Stele and Zakkur Stele, as well as m on the
Zakkur Stele.
189. 'Hadoram' is presumably short for 'Hadad-ram'.
190. This particular name is well attested among the corpus of Ammonite seals. See
further Aufrecht, A Corpus of Ammonite Inscriptions, p. 358.
191. Schniedewind, 'Tel Dan Stela', pp. 77, 80.
240 The Tel Dan Inscription
193. D.D. Luckenbill, Ancient Records of Assyria and Babylonia (New York:
Greenwood, 1927), pp. 245-46.
194. N. Na'aman, 'Hazael of Amqi and Hadadezer of Beth-rehob', UF21 (1995),
pp. 381-94.
195. Compare also V. Sasson, 'Murderers, Usurpers, or What? Hazael, Jehu, and the
Tell Dan Old Aramaic Inscription', UF28 (1996), pp. 547-54. Sasson takes the same
viewpoint as Na'aman.
242 The Tel Dan Inscription
Yet we must ask what purpose a multiple patronymic served the author
here. Kings are known to cite their own multiple patronymics,196 but not
that of an ally or enemy. This is limited to a single patronymic. Since the
king of Bayt-Dawid was apparently an enemy of the author,197 it is
unlikely that his multiple patronymic was cited by the author. Therefore,
we must have doubts about the validity of restoring the multiple patro-
nymic of Jehoash, son of Ahaziah, here, especially as it would require the
inclusion of at least three predecessors' names.
There is nothing within the inscription, however, which limits us to
identifying the personage in Line B7 as a figure from either Israel or Bayt-
Dawid. We may legitimately look outside these two entities to identify
him. A theophoric element preceding the element D1-, which could give
us a clue to the ethnicity of this person, is unfortunately no longer extant.
As such, we may propose a monarch of any nation in the Syria-Palestine
region.
The name U~h^ ("Ilram') is known to have been a popular name in
Ammon, perhaps even the name of a seventh-century king.198 The element
DT- is also to be seen in the name of the Edomite king, Aiarammu, known
from Sennacherib's annals (c. 701 BCE), and the popular Tyrian name,
Ahiram or Hiram. Noll mentions this latter possibility as a plausible option
for reconstructing this name in Line B7.199 On purely philological grounds,
he is correct. However, Noll dates the inscription along roughly the same
lines as Biran and Naveh. Since the present study has dated the inscription
to c. 800 BCE, it is unlikely that a king of Tyre was mentioned here, for at
that time Pygmalion ruled Tyre. The possibility still remains, though.
Noll's suggestion simply underlines the fact that the element D~l- was
common in names all over the Levant, thus leaving us no closer to finding
a solution.
Becking mentions the possibility that the name in Line B7 may be that
of 'Adonleram, steward of the royal house in Hamath.' Adonleram's name
is known to us from the inscribed bricks from Hamath. The title given to
him is rD^D'H"1 [H]pD. Yet, although "3D is predominantly used to denote
a governor, the fact that 'Adonleram is called 'governor of the royal
house' means his position was that of a steward rather than a regional or
political-administrative governor. As such, his position and title were akin
to the title fT^n-b^'H^N ('who is over the house'), known from seals and
bullae from Palestine. One wonders, therefore, what relevance a household
official in Hamath had for the author of the Tel Dan Inscription. A connec-
tion between 'Adonleram and a figure named in Line B7 is possible, but
unlikely.
Perhaps the best candidate for the person mentioned in Line B7 is a
figure known from the cache of ceramic inscriptions found at Hazor.
Specifically, the sherd of interest comes from Building 14a in Area A
of the Hazor excavations. The sherd is stamped with the word mDIDQ1?
('belonging to Makbirram').200 It is dated stratigraphically to the early
eighth century, making it synchronous with the Tel Dan Inscription. Al-
though no patronymic is given for Makbirram on the sherd bearing his
name, we may assume he was a wealthy and important figure in Hazor,
perhaps an administrative governor, a military general, a vassal ruler or
even a member of a royal family. The immense strategic significance of
Hazor to the regional balance also underlies the importance we may attach
to Makbirram, if indeed we can see him as a man of some authority. The
proximity of Hazor to Dan (less than 30 km) also makes him an especially
attractive prospect as a match for the personage in Line B7.201
However, a fair degree of uncertainty must be maintained because of the
fragmentary nature of the context and the relative ambiguity surrounding
Makbirram's person. Nevertheless, it must be said that Makbirram pre-
sents as good a match as any for the name in Line B7 in terms of the epi-
graphical, archaeological and historical data. Thus, the name in Line B7
may tentatively be identified with Makbirram of Hazor. It should, how-
ever, be realized that the choice of Makbirram as the most likely candidate
comes in the absence of any other candidate known from this era. As a
result, we simply do not know for certain who this figure in Line B7
actually was.
Line B8
"O1iT. Although the first letter of Line B8, a partially damaged yodh, lies
a fair distance from the following letter, a partially preserved heh, we must
consider them as part of the same lexeme. This is due to the fact that the
letters 111 do not make a meaningful lexeme on their own. Official Aramaic
transcribes the S.masc.sg. independent personal pronoun as in, but this is
not the case in Old Aramaic, where the pronoun is written KH. Hence it is
not the case here in the Tel Dan Inscription.
The only choice we have is to join the first three extant letters to form
the element 1iT. The fact that this is followed by the lexeme "ID ('son')
means we must view 1IT as a theophoric element within a name, rather
than as the independently standing divine name, Yahu. As the Mesha Stele
shows, this divine name preserves a heh during this period and is thus
transcribed as i"PiT. The apocopated form, 1IT, appears as the theophoric
element of numerous names originating in Israel and Judah where Yah-
wistic cults prevailed. The implication for Line B8 is that the figure men-
tioned here is almost definitely an Israelite or a Judaean.
For the century between 852 BCE and 753 BCE, no king of Israel pos
sessed a name with the final element being the theophoric element, "IIT.
Rather, three kings during this period (Joram, Jehoahaz and Jehoash) each
had Yahwistic names, but with the theophoric element as the first rather
than the last element of the name. From the biblical record, however, we
know that Amaziah became king of Judah in 796 BCE. It is, therefore, most
likely that Line B8 preserves the name [2V-]"l><iril'[:aiDK] ('Amaziah
ben-Joash').202 Once again, though, this is only a probability and not a cer-
tainty. We must factor in the possibility that this personage was not a
monarch. Amaziah ben-Joash, however, is quite likely considering the
time-frame in which we can date the original inscription. Hence, this recon-
struction of his name here is the most likely option, seeing that royal per-
sonages are likely to be named in such monumental lapidary inscriptions.
202. It must be remembered that there are two variant spelling of the name Joash/
Jehoash. For the sake of clarity with the Israelite king, Jehoash, we have chosen to label
the name of Amaziah's father as CW ('Joash'). However, it is quite possible that in the
Tel Dan Inscription, Joash's name was written as SDNliT rather than 2JNV. Since the bib-
lical text regularly interchanges the spelling of this name, there must be some ambiguity
as to how exactly to reconstruct the name here in the Tel Dan Inscription.
6. Textual Analysis 245
203. Although this word is only partially preserved, there is no doubt that this word
is a noun representing a numeral.
246 The Tel Dan Inscription
to suggest that the other parts of the original inscription did not use a
definite article. In fact, all the indications from the language of the text
lead us to believe that the post-positional definite article, N-, would have
been used as a normal feature of the language, Old Aramaic, or, if not ex-
pressly written, probably pronounced nonetheless. The text of the extant
fragments does not require the use of a definite article in these particular
portions of the text. Thus, the absence of the definite article from the
three fragments is circumstantial, not linguistic. Although the Aramaic of
Hazael's Horse Ornament and of the Sam'alian inscriptions preserve no
definite article, we cannot make the same assumption of the Tel Dan
Inscription, because we have no portion of text which requires the definite
article to be used. The close similarity of the Tel Dan Inscription's dialect
to that of the Sefire Treaties leads us to believe that the definite article was
most likely used in the Tel Dan Inscription. It is only the fragmentary
nature of the text that prevents us from seeing any explicit evidence.
Grammatical Survey
Orthography
Graphemes. A total of 21 different graphemes are attested on all three
fragments as a whole. The only grapheme that remains unattested is teth.
The single occurrence ofsadhe is also quite damaged, though there is no
doubt as to its identification.
Consonants. The text and its orthography indicate that there were likely
more consonantal phonemes than individual graphemes. By necessity,
therefore, some graphemes in the Tel Dan Inscription carry more than one
phonetic value:
1. The grapheme qoph (p) represents the phonemes q and d.
a. The representation of phoneme q is demonstrated by the
words Dip (Line A4), 'Dip (Line A5), bnp1 (Line A6),
[nn]^npl (Line A8) and p2K[ft] (Line B5).
b. The representation of phoneme d is demonstrated by the
words p-IKD (Line A4) and p"!K (Line A10).
2. The grapheme sin (V) represents the phonemes s and s.
a. The representation of phoneme s is demonstrated by the
words H^h (Line Al) and [btri]er (Lines A8, A12).
b. The representation of phoneme s is demonstrated by the
words DDEn (Line A3), 2TIS (Line A7), DEW (Line A9)
6. Textual Analysis 247
and in& (Line B5). The grapheme sin (527) may also have
served to represent the phoneme/, but such an instance is
unattested among the three extant fragments.
Vowels. The orthography of the Tel Dan Inscription represents only those
long vowels that occur in an open syllable at the end of a word. This is
demonstrated by the following cases:
1. At the end of a word, the grapheme heh can represent either the
vowel a or the vowel e in an open syllable.
a. The representation of vowel a in a final open syllable is
seen in the word HHK (Line A5).
b. The representation of vowel em a final open syllable is
seen in the word Hftn^nfrQ] (Line B2). The final heh of
this lexeme represents the 3.masc.sg. pronominal suffix.
It is, therefore, more indicative of the pronominal suffix
than of the accompanying vowel. Yet, it is to be expected
that other words ending in heh but not representing a pro-
nominal suffix did actually represent the vowel e. How-
ever, no such words are attested among the small portion
of text preserved on our three fragments.
2. At the end of a word, the grapheme waw represents a long 'u'-
type vowel in an open syllable.
a. The representation of long vowel u in a final open sylla-
ble is seen in the word in" [UEN] (Line B8). We might ex-
pect that the final vowel of the afformative for 3.masc.pl.
verbs was also represented by the grapheme waw and
indicative of long vowel u. However, no such verbs are
attested among the three extant fragments.
b. The representation of long vowel 6 is unattested among
the three extant fragments.
3. At the end of a word, the grapheme yodh can represent either the
vowel e (or a) or i in an open syllable.
a. The representation of vowel e in a final open syllable is
seen in the words ^[l]D[nm] (Line A2), 'S^M (Line A7)
and "HDN (Line B6). We might expect that this vowel was
properly represented by the grapheme heh (see 1 .a. above).
When the vowel e is represented by the grapheme yodh, it
is solely indicative of the masculine plural construct form.
Alternatively, if these sounds had not reduced to simple
248 The Tel Dan Inscription
Matres Lectionis. The orthography of the text indicates that matres lec-
tionis were not used within a word to indicate either a long or a short
vowel. Only long vowels in final open syllables were marked by certain
graphemes according to the vowel and grammatical circumstance (see the
Pronouns. The extant portions of the inscription attest the following pro-
nouns.
Independent personal pronouns. Only two independent personal pronouns
are attested among the extant fragments:
a. The l.com.sg. form, HDN (Line A5).
b. The 3.masc.pl. form, DPI (Line A10). This particular pronoun is a
special case since it is treated syntactically as a pronominal suf-
fix. Orthographically, however, it stands independently of the pre-
ceding noun.
Suffixed personal pronouns. Only two pronominal suffixes are attested
among the extant fragments:
1. S.masc.sg. (H-):
a. Attested only after verbal forms: nQn^hfrQ] (Line B2).
2. l.com.sg. C1-):
a. After nominal forms: ''UK (Lines A2, A3), *lhfo (Line
A6), ^Q (Line B3).
b. After prepositions: ""Dip (Line A5).
c. After object marker: [TP]K (Line B4).
Otherpronouns. Other than the two types of personal pronouns mentioned
above, no other types of pronouns are attested among the extant fragments.
Thus, there are no occurrences of demonstrative pronouns, interrogative
pronouns, or relative pronouns.
Nouns. The inscription contains numerous nouns, many of which are un-
damaged. Some, however, require reconstruction, while others are identi-
fiable as nouns but defy specific reconstruction.
Forms. The following types of nominal forms are attested among the three
extant fragments:207
Table 6.1. 'Paradigm' of masculine nouns, based on the information from the
Tel Dan Inscription.
Masculine Forms
Singular Plural
Absolute CHS (Line A7) ]'[...] (Line B6)
*)^np
Construct I^D (Line A8) **b* (Line A7)
Emphatic *8^np *K^np
been removed when the noun is quoted. As such, only the singular absolute forms are
quoted.
6. Textual Analysis 251
Verbs. There are five verbal stems attested on the Tel Dan fragments:
1. Pe'al
2. Pe'il.
3. Pa'el.
4. Haph 'el.
5. Ithpe'el.
Within these, we may recognize the following categories of aspect:
252 The Tel Dan Inscription
1. Perfect.
2. Imperfect.
3. Infinitive.
4. Jussive.
5. Participle.
These are the conjugations and aspects attested in the fragments. There
is no evidence of, nor any need to see, the presence of consecutive imper-
fects derived from a preterital form.
The Ithpe 'el is attested here only in the infinitive within a damaged
context (Line B2). It is, therefore, difficult to ascertain whether the pre-
formative of the stem in the perfect and/or infinitive conjugations was
formed with a -Tin particle or an -TIN particle. An Ithpa 'al perfect form is
attested in Bar Rakib A. 14 with a -Till particle CDtWnn), suggesting that a
-Tin particle is plausible for an Ithpe 'el form. All other attested forms of
Ithpe 'el or Ithpa 'al are in the imperfect conjugation, some being ambigu-
ous due to damaged contexts.2081 have opted for a -fin particle on the
basis of the attestation of the Ithpa 'al perfect form in Bar Rakib.
Names of Places
mmi 'Bayt-Dawid', n.pr.loc. [name of the city state centred on Jerusalem]-
Line A9.
^tner 'Israel', n.pr.gent.Line A8; [^n]2T, Line A12.
]"1DE? 'Samaria', n.pr.loc.[flDEJ], Line A13.
Chapter 7
HISTORICAL COMMENTARY
Introductory Remarks
A synthesis of the amassed data from the Tel Dan Inscription now puts us
in a position to extract historical implications from the inscription. As with
all ancient Near Eastern history, such reconstruction must proceed with a
fair degree of caution. The corroboration of various kinds of sources is no
easy task, particularly when their milieu is so far removed from our own
time and culture.
With the Tel Dan Inscription we have a primary written source. Further-
more, information can be gleaned not just from its text, but also from its
physical remains. The fact that the inscription is a primary source is of
great importance because it lends weight to the information we can extract
directly from it. Our caveat must be, however, not to take all the infor-
mation derived from the inscription at face value, nor to be swept up in an
attempt to impose the inscription onto a neat biblical template. Neither
must we dismiss the biblical evidence from having any value for historical
reconstruction. Rather, we must recognize that the biblical texts are sec-
ondary sources and employ them with the appropriate contingent weight.
Historical Considerations
The Author of the Inscription
Although the author of the Tel Dan Inscription is never named in any of
the three extant fragments, we may look to the clues given in the text,
language and archaeological context of the inscription in order to attempt
an identification. Since the inscription belongs to the latter period of the
second construction phase at Dan's city gates, our author lived sometime
c. 800 BCE. During this time, the city of Dan was under Aramaean influ-
ence. Under the leadership of Hazael, Damascus had grown to be the head
of a quasi-empire with influence over a number of Levantine states. Dan,
256 The Tel Dan Inscription
Lines Al 1-12 refer to a ruler of Israel in the third person. This comes
not long after Line A8 where the author makes mention of having killed a
king. Since the mention of the king of Israel is in the immediate context,
and Line A12 mentions the rise of a new Israelite king, it is clear that the
author was not the king of Israel. Together with the fact that the nearby
Aramaean state of Damascus was at its zenith towards the end of the ninth
century BCE, we should look to monarchs of that state for the author of the
Tel Dan Inscription.
Most theories about the Tel Dan Inscription attribute authorship to
Hazael. However, such claims are based on a reading derived from the old
arrangement of the fragments, now seen to be defunct. This reading identi-
fied the two names in Lines B7 and B8 as those of '[Jehojram son of
[Ahab]' and '[Ahaz]iah son of [Jehoram]', respectively. Since the death of
these two kings occurred in c. 841 BCE, and the author could not have been
the king of Israel, scholars nominated Hazael as the only logical choice for
the author of the Aramaic inscription. The redundancy of the old fragment
arrangement and the more detailed archaeological picture we now have
from Tel Dan means we must rethink this entire hypothesis.
Data gleaned from the archaeological context of the fragments and the
palaeography of the script indicated that the inscription was broken at the
beginning of the eighth century BCE and probably written not very many
years before that. In fact, the excellent condition of the inscribed letters, as
well as the dating of the script, suggests that the original stele was not
standing for many years at all before it was knocked down and fractured.
Thus, we must look to the period at the very end of the ninth century BCE
and the beginning of the eighth century BCE for the author. During this
time, Aram-Damascus was the dominant power in Syria-Palestine. Under
the leadership of Hazael, Aram-Damascus had gone from a small state on
the fringe of the Syrian steppes and desert, to a quasi-empire with influ-
ence over at least the majority of states in Syria-Palestine.8 Hazael's death
occurred in c. 799 BCE9the very time to which the archaeological and
palaeographical evidence point for the date of the inscription's production
and destruction. We must, therefore, nominate both Hazael and his son,
Bar Hadad, who succeeded him on the Damascene throne, as the possible
authors of the Tel Dan Inscription.
14. B. Margalit, 'The Old-Aramaic Inscription of Hazael from Dan', UF26 (1994),
pp. 317-20(317).
15. Margalit, 'The Old-Aramaic Inscription', p. 318.
16. Yamada, 'Aram-Israel Relations', pp. 613-14.
7'. Historical Commentary 261
Hazael could have been of royal blood but not necessarily in line of succes-
sion to the throne of Aram-Damascus. Considering that in biblical times
frequently a monarch had several wives and various heirs, it is possible that
Ben Hadad was Hazael's father.17
27. This situation compares with Hazael's ability to recoup after the campaigns of
Shalmaneser III against him in 841 and 838 BCE.
28. See Zakkur B3-15.
29. The precise dates for Hadianu's reign cannot be calculated. As mentioned above,
he had become king of Aram-Damascus by 773 BCE. His successor, Resin (Akkadian
Rahianu), is mentioned by Tiglath-Pileser III as a contemporary of Menahem of Israel.
Menahem died in 741 BCE. Thus, by this time, Hadianu had died. It is likely that Resin
had already succeeded to the throne before the coronation of Tiglath-Pileser III in 745
BCE, putting Hadianu's death sometime before that.
266 The Tel Dan Inscription
Hazael was arguably the most influential and successful monarch of Aram-
Damascus. His memory is even recounted in the oracles of Amos who
prophesied at Bethel nearly half a century after Hazael's death.30 Both the
Hazael Ivory and the Horse Ornament from Samos both refer to Hazael as
]N~1Q ('our lord')- It appears that this title passed to his son, Bar Hadad II,
and it was this title that was used of Bar Hadad II in the Tell el-Rimah
SteleMari.
At the time of Hazael's death, Aram-Damascus had taken control of the
region's most important trade routes. From Damascus, the vital highway
to Tadmor (Palmyra) and Mesopotamia branched to the northeast, while
the road to Hamath, Aleppo and northern Syria branched to the north. Bar
Hadad's later campaign against Zakkur of Hamath and Luath may well
have been partly motivated by a desire to restore the territory conquered
by his father. Numerous biblical texts also inform us of Hazael's capture
of Transjordanian territories, giving him the King's Highway and sub-
sequently the rich trade from Arabia. Amos's oracles condemn Damascus
for a violent conquest of Gilead (Amos 1.3-5), and Israel's boast of having
captured Lo-Debar and Qarnayim (Amos 6.13) implies that these Transjor-
danian towns previously had been in the secure possession of another
state. This can only have been Aram-Damascus. This tallies with 2 Kgs
10.32-33, which has Hazael extending his control all the way to the Arnon
Gorge. If this had indeed been the case, then Hazael would also have
controlled the caravan route through the desert between Rabbath-Ammon
and Dumah in the northern Arabian Desert.
The biblical record also has Hazael campaigning against Gath and
seizing it (2 Kgs 12.17-18). Certainly, the downfall of Gath alluded to in
Amos 6.2 was legendary. There is a question, however, as to whether this
description refers to Gath's conquest at the hands of Hazael (late ninth
century BCE) or the hands of Sargon II (712 BCE).31 If it refers to the latter,
then we must date the reference in Amos to a period well after Amos's
ministry (c. 760 BCE). Since Gath's final downfall seems to have come
with Sargon II in 712 BCE (or with Sennacherib in 701 BCE at the latest), it
is likely that the description of Gath's downfall as legendary derives from
that time.32 Although Gath may simply have been known for its many
downfalls, further support for this late date is the fact that the downfall of
30. Amos's oracles can be dated between c. 760-750 BCE. The reference to 'the
house of Hazael' is seen in Amos 1.4.
31. SeeANET,p. 286.
32. Compare Mic. 1.10.
7. Historical Commentary 267
Kalneh and Hamath are also described in Amos 6.2 as comparable to that
of Gam. Both these cities fell to Tiglath-Pileser III in 738 BCE. It is possi-
ble that Amos 6.2 refers to the conquests of Kalneh, Hamath and Gath by
Hazael, but this is much less likely.
Nevertheless, it is not inconceivable that Hazael did indeed campaign on
the Philistine Plain. Israel was certainly weak enough politically, economi-
cally and militarily for this to have occurred. The tradition of a harsh op-
pression of Israel by Damascus, particularly at the hands of Hazael, is
quite pervasive33 and would be in line with the geopolitical situation of the
late ninth century BCE. Recent excavations at Tell es-Safi, the most likely
location of Gath, have unearthed a destruction layer which has been dated
stratigraphically, typologically and radiometrically to the late ninth century
BCE or early eighth century BCE.34 Whether this destruction can be attrib-
uted to Hazael is debatable, though it is plausible. Similar destruction
levels dated to the same time have also been found at Tell er-Rumeilah
(Beth Shemesh) and Tell Hamid (Gittaim?). If the debris from these sites
can be attributed to Hazael, then it would appear that the Damascene
monarch conducted a campaign against the Shephelah region. However, it
is more likely that the debris levels from these sites are connected to either
the expansion of the Judaean state in the first half of the eighth century
BCE or, more likely, the earthquake that struck the region in c. 760 BCE.35
This still does not preclude Hazael from having campaigned at Gath.
For such a campaign to have been conducted, Hazael must have had some
kind of control over the routes in the Jezreel Valley, the Aruna Pass and
the Sharon Plain. A successful campaign against Gath would then have
given him control over the coastal road (the 'Way of the Sea') between
Egypt and Palestine. 2 Kings 12.17-18 also has Hazael initiating a cam-
paign towards Jerusalem. If this indeed occurred, it was presumably to
affect control over the road along the watershed of Palestine's hill country,
which connected with the Road to Shur and the Road to the Arabah in the
Negev. Although the biblical account has Hazael turning away from Jeru-
salem, it implies that he was able to gain effective control of these routes
through the submission of Joash of Judah. If Hazael had actually cam-
paigned that far south, his control of these regions must have been nomi-
nal or indirect, purely because of the seasonal economy of Damascus (as
during this construction phase. This leaves Jehoahaz who died in 798 BCE.
This date for Jehoahaz's death fits perfectly with the archaeological and
palaeographical data of the Tel Dan Inscription. Therefore, we can con-
clude that Jehoahaz ben-Jehu was the king of Israel whom Bar Hadad
claims to have killed in Line A8.
In Lines Al 1-12, Bar Hadad goes on to mention the successor of the
Israelite king whom he had killed. The biblical record states that Jehoash
ben-Jehoahaz came to the Israelite throne after his father's death (2 Kgs
13.9). This is corroborated by the Tell er-Rimah Stele, which records that
Adad-Nirari III received the tribute of 'lasu' (Joash) of Samaria.36 This
must have occurred during 796 BCE when Adad-Nirari III campaigned
against Mansuate in the Lebanese Beqa', just two years after Jehoash's
accession to the Israelite throne in Samaria. Hence, I have restored the
name 'Jehoash' or 'Joash' to the end of Line Al 1.
The biblical record of Jehoahaz's reign is a predominantly theological
statement. However, it describes a political situation that seems to be well
in keeping with the geopolitical circumstances of the timethat is, con-
flict between the states of Aram-Damascus and Israel:
In the twenty-third year of Joash ben-Ahaziah, king of Judah, Jehoahaz
ben-Jehu reigned over Israel in Samaria for seventeen years. Yet, he did
evil in Yahweh's eyes and followed the offences of Jeroboam ben-Nebat,
who had made Israel offend. He did not stray from them. So Yahweh's
anger burned against Israel and he put them into the hand of Hazael, king of
Aram, and the hand of Bar Hadad ben-Hazael all the time. (2 Kgs 13.1-3)
This account of conflict between the monarchs of the two Levantine
states squares well with Bar Hadad's own account of the conflict in the Tel
Dan Inscription. However, what the biblical record does not relate is that
Jehoahaz was killed by his nemesis, Bar Hadad. This piece of information
is glossed over by the following statement:
Yet Jehoahaz beseeched Yahweh's person and Yahweh listened to him
because he saw Israel's oppression. Indeed, the king of Aram had oppressed
them. Yahweh gave Israel a saviour and they escaped from the hand of
Aram. Then the Sons of Israel lived in their own tents as in times past.
(2 Kgs 13.4-5)
36. In the biblical texts, the names 'Jehoash' and 'Joash' are interchangeable. The
former must be considered the proper full name and the latter a hypocoristicon.
270 The Tel Dan Inscription
forced to pay a hefty tribute to the Assyrian king. The Saba'a Stele puts
the tribute at 100 talents of gold, 1000 talents of silver, as well as other
sums that have been damaged on the face of the stele.37 The Tell er-Rimah
Stele puts the tribute payment at 2000 talents of silver, 1000 talents of cop-
per, 2000 talents of iron and 3000 multicoloured and plain linen gar-
ments.38 The Nimrud Slab puts the tribute at 2300 talents of silver, 20
talents of gold, 3000 talents of copper, 5000 talents of iron, coloured wool-
len and linen garments, an ivory bed, a couch inlaid and embossed with
ivory, and an immeasurable amount of Bar Hadad's property and goods.39
Whichever account is relied upon, it is abundantly clear that Bar Hadad
paid an enormous tribute to the Assyrian king. The Tell er-Rimah Stele
also tells us that the king of Israel paid an undisclosed tribute to Adad-
Nirari, though this king was not Jehoahaz but his son, Jehoash. Thus, the
notion that Israel's reprieve from Aram-Damascus came in Jehoahaz's life-
time, should be altered. The Tell er-Rimah Stele and the Eponym Chron-
icle show that the reprieve came during the reign of his son, Jehoash.
Bar Hadad had inherited from his father, Hazael, a large sphere of influ-
ence in Syria-Palestine. After Adad-Nirari Ill's castigation of Bar Hadad
in 796 BCE, Aram-Damascus began to lose sway over the region. The
archaeological record of Dan shows that at this precise time, a new phase
of construction occurred at the extremities of the southern city gate. This
can be associated with Dan's incorporation into Israel. The biblical record
corroborates this:
Hazael, king of Aram, died and Bar Hadad, his son, reigned in his stead.
Then Jehoash ben-Jehoahaz went back and took from the hand of Bar
Hadad ben-Hazael the towns that he had taken from the hand of Jehoahaz,
his father, in war. Jehoash had defeated him three times when he retrieved
the towns of Israel. (2 Kgs 13.24-25)
This apparent alliance between the king of Israel and the king of Bayt-
Dawid, as well as Bar Hadad's reference to 'their land', must mean that
the region or town of Bayt-Dawid was an immediate neighbour to the king-
dom of Israel. Presumably, this means the two states shared a common
border and a single terrain. This information places Bayt-Dawid some-
where in the central hill country of Palestine. Of this area, Israel occupied
the country encompassed by the Jezreel Valley, the Jordan River, the
Sharon Plain and the mountainous region just south of Bethel. This leaves
the southern hill country south of Bethel as the area in which we should
look for Bayt-Dawid. All the other frontiers of Israel can be classed as
another terrain and are, therefore, unlikely to have been considered a unity
with the territory of Israel.
As mentioned previously, no entity named 'Bayt-Dawid' has hitherto
been known to us. Either we are dealing with an entity previously known
under a different name or we have a new entity altogether. If Bayt-Dawid
was a city-state, we need to look among the individual sites to Israel's
immediate south which date at least to the end of the ninth century BCE.
The leading candidates for a city-state dating to this time are Gibeon (el-
Jib), Gezer (Tell Jezer), Jerusalem, Beth Shemesh (Tell er-Rumeilah),
Lachish (Tell ed-Duweir) and Hebron (er-Rumeidah). However, for most
of these cities, there is no evidence that they were called 'Bayt-Dawid' in
antiquity. Specifically, in the cases of Gibeon, Gezer, Lachish and Hebron,
we have evidence that these cities were known by the names by which we
know them also.40
Beth Shemesh is a particularly interesting possibility as the location of
Bayt-Dawid. Major installations for the production of olive oil and wine,
as well as minor installations for dyeing have been found at the site (Tell
er-Rumeilah) dating to Iron II.41 This suggests the importance of Beth
Shemesh as a leading centre for these industries, either in co-operation or
rivalry with nearby Eqron. Among the ceramic evidence, jar handles
stamped with the word "fbD1^ ('belonging to the king') have been found.
However, these stamped handles do not indicate that Beth Shemesh had a
resident king. Hundreds of similarly stamped jars have been found at
40. In addition to the usage of these names in the biblical texts, Lachish is named in
Assyrian texts, while the name of Hebron appears stamped on numerous storage jars
dating to the late eighth century BCE. See further Ahlstrom, The History of Ancient
Palestine, pp. 699-700.
41. E. Grant and G.E. Wright, A in Shems Excavations. V. Text and Plates (5 vols.;
Bible and Kindred Studies, 8; Haverford: Haverford College, 1939), pp. 75-77.
7. Historical Commentary 273
numerous sites throughout Judah.42 Many of them also bear the name of a
city, though only four cities are attested from these stamped jarnamely,
Hebron, Socoh, Ziph and the enigmatic H^QQ. It is quite easy to read these
stamps as construct expressions, such as 'Belonging to the king of Hebron'
or 'Belonging to the king of Socoh'.
However, numerous factors count against this interpretation. First, these
storage jars are all of the same essential type and the letters l^fcb are
accompanied by an image of either a winged sun-disk or a scarab beetle.
Certainly the scarab beetle image has been identified on two bullae as part
of the personal seal of Hezekiah.43 The iconography, palaeography, epi-
graphy and archaeological context of these jars point to a date towards the
end of the eighth century BCE.44 They also reflect a situation in which all of
Judah was under the central administration of Jerusalem. All the seal
impressions on these particular storage jars throughout Judah were also
made with a small number of seals.45 Also, the clay that the jars them-
selves are made of appears to have a provenance in or around Jerusalem
and the region around Mareshah and Moreshet-Gath in the Shephelah.46
The enigmatic name HO2Q has variously been interpreted as a reference to
'Jerusalem' or the government of Judah centred in Jerusalem.47 To this we
may add that it might be a reference to Mareshah or Moreshet-Gath (per-
haps a misspelling?). Whatever the interpretation of the name, however,
these jars were not made in the cities whose names are stamped on them
(unless a connection between HO2Q and Mareshah can be made). Rather,
they were distributed to numerous towns in Judah, perhaps with the
intention of transferring them to the cities stamped on the handles.48 Thus,
42. D. Ussishkin, 'Royal Judean Storage Jars and Private Seal Impressions',
BASOR 223 (1976), pp. 1-13; 'The Destruction of Lachish by Sennacherib and the
Dating of the Royal Judean Storage Jars', Tel Aviv 4 (1977), pp. 28-60.
43. P.M. Cross, 'King Hezekiah's Seal Bears Phoenician Imagery', BARev 25.2
(1999), pp. 42-45, 60.
44. Ussishkin, 'The Destruction of Lachish by Sennacherib', pp. 28-60; Cross,
'King Hezekiah's Seal Bears Phoenician Imagery', p. 44.
45. J.M. Miller and J.H. Hayes, A History of Ancient Israel and Judah (London:
SCM Press, 1986), pp. 354-56.
46. Ahlstrom, The History of Ancient Palestine, pp. 699-700.
47. H.L. Ginsberg, 'MMST andMSH', BASOR 109 (1948), pp. 20-21; Y. Aharoni,
The Land of the Bible: A Historical Geography (London: Westminster Press, 2nd rev.
edn, 1979), pp. 298-99.
48. For various theories about the purpose of these stamped jars and the signifi-
cance of the cities stamped on their handles, see N. Na'aman, 'Hezekiah's Fortified
274 The Tel Dan Inscription
at the end of the eighth century BCE, Beth Shemesh was not a city-state,
but part of the regional state of Judah.
Yet, was this also the situation earlier in that same century? In the
biblical record, Beth Shemesh is portrayed as the centre of a territorial
dispute between Israel and Judah during the reigns of Jehoash of Israel and
Amaziah of Judah (2 Kgs 14.8-14). The decided supremacy of Israel at the
beginning of the eighth century BCE is well attested by the archaeological
record. During the reigns of Jehoash and his son, Jeroboam II, Israel ex-
perienced a cultural and economic floruit. Judah at this time was a less
well established state, just emerging from veritable insignificance on the
international political arena, though on the rise as an organized state. It is
only to be expected that if Israel and Judah did indeed clash at this time,
as the biblical text purports, that Israel would have been the victor. The
question is whether Israel and Judah actually clashed. The answer to this
question is that they most likely did, especially since both states were
expanding during the reigns of Jehoash and Amaziah. Possession of the
agriculturally rich Shephelah, particularly the city of Beth Shemesh, would
undoubtedly have become an issue at some stage. There is no reason, then,
to dismiss the biblical text, which reports what is only to be expected
given the archaeological evidence.
This being the case, it is unlikely that Beth Shemesh was an independent
city-state in c. 800 BCE. It appears to have been a city of certain importance,
but not a state capital independent of either Israel or Judah. In excavations
at the site of Beth Shemesh (Tell er-Rumeilah), no palatial structures dating
to Iron II have been found. One large structure was uncovered which Grant
and Wright call a 'residency', but even by their own admission, there is
nothing about the structure to suggest it was a palace or even a governor's
residence.49 Furthermore, the wall of Beth Shemesh was severely breached
during the destruction of the Stratum IV city, probably at the hands of the
Philistines.50 These breaches were repaired with rubble during subsequent
Cities and the LMLK Stamps \BASOR 261 (1986), pp. 5-21; Y. Yadin, 'The Fourfold
Division of Judah', BASOR 163 (1961), pp. 6-12.
49. Grant and Wright, Ain Shems Excavations, V, pp. 69-71.
50. Grant and Wright date the destruction of Stratum IV to the end of the thirteenth
century BCE claiming that the Philistines were unlikely to have been the perpetrators;
see Grant and Wright, Ain Shems Excavations, V, p. 12. However, following Finkel-
stein's observations regarding the relation between Monochrome ware and sites from
the 20th Dynastysee Finkelstein, 'The Archaeology of the United Monarchy'I am
inclined to downdate Grant and Wright's conclusion by the better part of a century.
7. Historical Commentary 275
(701 BCE), Jerusalem had been strongly fortified. With the city expanding
into the 'Second Quarter' (i~I]O2n) on the Western Hill, a new 'broad wall'
was erected to protect these new outer environs.53 At this time, we can see
the influence of Jerusalem extending to all corners of Judah's greater area.
Among the evidence for this is Sennacherib's own annals recounting his
campaign in 701 BCE. Sennacherib mentions having besieged 46 of Heze-
kiah's cities and taking them from Hezekiah's country and giving them to
various Philistine kings.54 This must be a reference to a regional state of
Judah. Sennacherib also calls Hezekiah 'the Judaean' (Akkadian laudai),
obviously in reference to the country that Hezekiah ruled.
Prior to this time, however, Jerusalem was a smaller town, confined to
the ridge popularly known as the 'City of David' and the area roughly
equating to the current Temple Mount. Amos's oracles against the nations
(Amos 1.3-2.16) reflect a political situation in c. 760 BCE in which a state
of Judah did exist, with Jerusalem as the capital. Yet, how large was the
state of Judah at this time? Certainly Amos himself was considered a
Judaean with his home in Tekoa.55 Thus, we are safe in determining that in
c. 760 BCE the state of Judah extended from the northern environs of Jeru-
salem along the Central Ridge of Palestine to at least Tekoa, some 15 km
south of Jerusalem. If the accounts of the reigns of Amaziah and Uzziah
(Azariah), kings of Judah, are accurate, we have a situation in which
Judah's influence extended all the way to Edom,56 even as far as the Gulf
of Aqabah.57 This is indeed a plausible situation considering that the for-
mer power of the region, Aram-Damascus, had been considerably weak-
ened and its territories along the King's Highway taken from its control.
Judah could easily have taken advantage of this new situation to pursue its
53. The 'Second Quarter' (TUBOn) is mentioned in 2 Kgs 22.14 (= 2 Chron. 34.22).
The building of this new outer wall is recalled in 2 Chron. 32.5 and probably also in
Isa. 22.10-11. This must be the 'Broad Wall' uncovered by Avigad. See N. Avigad,
'Excavations in the Jewish Quarter of the Old City of Jerusalem, 1969/1970', 7E/20
(1970), pp. 1-8, 129-40; idem, 'Excavations in the Jewish Quarter of the Old City of
Jerusalem, 1971', IEJ22 (1972), pp. 193-200; A.D. Tushingham, 'The Western Hill
under the Monarchy', ZDPV95 (1979), pp. 39-55. For the possibility that this outer
wall was built after Sennacherib's campaign in 701 BCE, see N. Avigad, Discovering
Jerusalem (Nashville: Thomas Nelson, 1983), pp. 46-57.
54. See ANET, pp. 287-88.
55. This impression is gained by the superscription to Amos's oracles (Amos 1.1)
and the record of his encounter with Amaziah, priest of Bethel (Amos 7.12).
56. Compare 2 Kgs 14.7, 10.
57. Compare 2 Kgs 14.22.
7. Historical Commentary 277
58. For the course of these and other relevant roads, see Y. Aharoni and M. Avi-
Yonah, Macmillan Bible Atlas (New York: Macmillan, 3rd rev. edn, 1993), p. 17, Map
10.
59. This is evidenced by a notable shift in ceramic culture. See O. Zimhoni, Studies
in the Iron Age Pottery of Israel: Typological, Archaeological, and Chronological
Aspects (Tel Aviv: Institute of Archaeology, Tel Aviv University, 1997), pp. 172-73.
60. The exact chronology for Strata IV and V at Lachish is still not fixed. However,
the end of Lachish IV can be placed in the early eighth century BCE. Its cultural con-
tinuity with the preceding Lachish V leads us to consider Lachish IV as representing at
least the second half of the ninth century BCE. See further Zimhoni, Studies in the Iron
Age Pottery of Israel, pp. 172-74.
278 The Tel Dan Inscription
along with Ammon and Moab, Judah was a particularly weak and
politically insignificant region. Indeed, Ammon and Moab must have
suffered considerably while Aram-Damascus controlled the King's High-
way. If Hazael did exact tribute from Joash of Judah,63 then Judah must
have been put into a similarly weak state.
It is with this poor, insignificant and fractured region that we should
associate the toponym Bayt-Dawid. This locality was essentially confined
to the southern spur of Jerusalem, though it may have had some influence
over a few neighbouring hills and valleys. After the campaign of Sennach-
erib against Hezekiah in c. 701 BCE, Judah was again essentially reduced
to these limitsa fact reflected in Esarhaddon's annals, where Hezekiah's
successor, Manasseh, is called sar URUIa 'udi ('king of the city of Judah').64
The connection between Bayt-Dawid and Jerusalem is clinched when we
understand the nature of Jerusalem prior to the economic boom of the
eighth century BCE and the names associated with the town.
In c. 800 BCE the city of Jerusalem was confined to the narrow spur
between the Qidron Valley and the Central (Tyropoean) Valley. We can
also presume the presence of a temple on the Temple Mount. A fortifying
wall connected these two sections of the city in the ninth century BCE at
the earliest.65 The Western Hill to the west of the Central (Tyropoean)
Valley was not incorporated into the city until well into the reign of Heze-
kiah.66 That there was previous settlement on the Western Hill is demon-
strated by the fact that Hezekiah's wall enclosing this 'Second Quarter'
was built over some of the ruins of previously standing houses. However,
it is unlikely that there was much, if any, settlement on the Western Hill in
c. 800 BCE. The ceramic wares taken from these demolished houses date
from no earlier than the eighth century BCE,67 pointing to settlement of the
Western Hill sometime during that century rather than during the previous
ninth century BCE.
Thus, the city of Jerusalem in c. 800 BCE occupied only the spur known
as the 'City of David' or 'Ophel'. The biblical text knows of this term,
"111 TU ('City of David'), as an early or alternate name for Jerusalem. The
name is most frequently associated with the place of burial for the kings of
Judah. Up until the time of Jerusalem's expansion under Hezekiah, the bib-
lical text places the burial of every Judaean king in the 'City of David'.68
Importantly, in 2 Sam. 5.7 it is associated with the 'fortress of Zion', cap-
tured by and renamed after David. It appears, therefore, that the name "IT
"111 ('City of David') is associated with references to the early stages of
Jerusalem-namely, the period before the end of the eighth century BCE. It
has a strong tradition of association with a fortified compound and its
royal owners.
It is my contention that the toponym 'Bayt-Dawid' (TITTD), as used in
the Tel Dan Inscription, is the equivalent of the biblical 'City of David'
("111 TU). The nounTU ('city') is unattested in Old Aramaic, so there is
no surprise that this element has been substituted with -TPD in the Old
Aramaic of the Tel Dan Inscription. The reason why the noun iT"!p was
not used instead may be due to the fact that Jerusalem, at this time, was
more of a citadel or castle compound rather than an urban settlement.69
Nevertheless, we do see an interchange between the element -TU and -TVI3
when, in Josh. 19.41, the town of Beth Shemesh is referred to as VlftV} "IT
('Ir Shemesh'). Thus, the interchange is certainly not unheard of.
Since Bayt-Dawid is the equivalent of 'City of David', we must there-
fore understand the toponym 'Bayt-Dawid' as a reference to a landed
estate or city-state. It should not be regarded as a reference to a regional
state, for had the City of David stood as a capital of a regional state, then
we would expect Bar Hadad to have referred to this state by its regional
title, rniiT ('Judah'). Since this is not the case, 'Bayt-Dawid' should be
considered a landed estate or a city-state that lent its name to the immedi-
ately surrounding district, rather than to the entire region of Judah. Thus,
the Tel Dan Inscription confirms that Jerusalem's sovereignty over the
regions of Judah was more token and ideological than an actual political
reality. From the point of view of the region's most dominant power of the
68. There is some confusion over the precise of location of Ahaz's tomb in the
biblical accounts. 2 Kgs 16.20 places his tomb in the City of David, while 2 Chron.
28.27 sees Ahaz buried 'in the city, in Jerusalem', though not in the same tombs as the
kings of Israel. Despite this odd statement about the precise location of Ahaz's tomb, it
appears clear that he was buried in the City of David. The text of Kings fails to men-
tion the location of Hezekiah's tomb. The Chronicler, however, tells us that he was
buried in the 'upper tombs of David's sons' (2 Chron. 32.33). This was probably
located in the City of David.
69. Compare Sabean "1U meaning 'fortified height'.
7'. Historical Commentary 281
70. Mazar and Mazar, Excavations in the South of the Temple Mount, pp. 13, 29.
71. E.g. A. Jepsen, 'Israel und Damaskus', AfO 14 (1941-45), pp. 153-72; C.F.
Whitley, 'The Deuteronomic Presentation of the House of Omri', VT 1 (1952),
pp. 137-52; J.M. Miller, 'The Elisha Cycle and the Accounts of the Omride Wars', JBL
282 The Tel Dan Inscription
85 (1966), pp. 441-54; idem, 'The Fall of the House of Ahab', VTll (1967), pp. 307-
324; idem, 'The Rest of the Acts of Jehoahaz', ZAW 80 (1968), pp. 337-42; E.
Lipinski, 'Le Ben-Hadad II de la Bible et 1'histoire', in P. Peli (ed.), Proceedings of the
Fifth World Congress of Jewish Studies (4 vols.; Jerusalem: World Union of Jewish
Studies, 1969), I, pp. 157'-Ibidem,' Aram et Israel duX e auvm e siecleav.N.E.',v4c/a
Antiquita 27 (1979), pp. 49-102 (75-76); Pitard, Ancient Damascus, pp. 114-25;
Yamada, 'Aram-Israel Relations', pp. 615-18.
72. Although the numbers of the southern coalition were probably inflated in the
Assyrian texts, Ahab's importance and strength within the coalition is not undermined.
The fact that he could send a chariot and infantry force to Qarqar in northern Syria that
was worthy of mention by the Assyrians confirms this fact.
7. Historical Commentary 283
further demonstrates the fact that these two Levantine kings were clearly
rivals, not allies.73
Thus, it appears that until Hazael's seizing of the Damascene throne,
there was an active entente between Israel and Aram-Damascus. It is, there-
fore, odd that Ahab, a clearly willing member of this entente cordiale, is
portrayed in the biblical text as constantly embroiled in hostilities with his
Damascene counterpart.74 This does not seem to comport with the political
situation of the time. With regard to this discrepancy, three biblical texts
are of interest to us.
1 Kings 22.1-40. The first episode we need to deal with is that of Ahab's
death. The narrative recounts a campaign in which Ahab of Israel and
Jehoshaphat of Judah join forces to fight the Aramaeans and recapture
Ramoth-Gilead. This was a Transjordanian city on the King's Highway
which the Aramaeans were in possession of, but to which Israel had tradi-
tional claims. In the narrative, the king of Israel's tactic is to head into the
battle incognito, but he suffers a blow from an arrow shot at random.
Watching the battle from his chariot, the king of Israel succumbs to his
wound and dies. Upon his death, his forces retreat and are defeated. The
king's body is then conveyed to Samaria and buried there.
In this narrative, the king of Israel remains largely anonymous. It would
appear that the names of both the king of Israel and the king of Judah, when
they are expressly named in the narrative, have been supplied at a much
later stage of the narrative's transmission in order to discredit the Omride
Dynasty, especially Ahab. The geopolitical situation at the end of Ahab's
reign does not permit such a battle between Israel and Aram-Damascus to
have been fought for possession of Ramoth-Gilead. In the year of Ahab's
death, 853 BCE, both Ahab and Hadad-'idri of Aram-Damascus had been
allies in the anti-Assyrian coalition that confronted Shalmaneser III at
Qarqar and checked his advance.
Lipinski suggested that this narrative of Ahab's death actually recounts
the death of his son, Jehoram.75 However, since Jehoram was assassinated
in Jehu's coup, this theory does not fit the evidence. The narrative does,
however, fit the time of Jehoahaz remarkably well, especially considering
the information gleaned from the Tel Dan Inscription. Dijkstra suggested
the plausibility of this scenario in connection with the Tel Dan Inscription,
but rejected it on the grounds that the archaeological context did not permit
a date for Fragment A at the turn of the eighth century BCE.76 Dijkstra's
article, however, appeared before the discovery of Fragments Bl and B2.
Furthermore, the evidence from excavations subsequent to the discovery of
these fragments proves that the original inscription must date to the turn of
the eighth century BCE. Therefore, the connection with the reign of Jehoahaz
is immediately apparent and applicable to the biblical text.
During Jehoahaz's lifetime, Hazael had taken possession of Israel's
Transjordanian territories. Understanding 1 Kgs 22.1-40 in connection
with Jehoahaz, it appears that upon Hazael's death, Jehoahaz tried to renew
resistance against Aramaean domination and attempted to retake Ramoth-
Gilead in Transjordan. To do this he enlisted the help of Joash, king of
Bayt-Dawid, and went into battle against Hazael's successor, Bar Hadad
II. In this battle, Jehoahaz lost his life and Israel was presumably defeated.
Thus, the biblical account of Ahab's death has actually been derived
from the memory of Jehoahaz's death at the hands of Bar Hadad. The
memory of this event is still preserved in the Tel Dan Inscription. Just how
Ahab actually died remains an unanswered question. It has been posited
that Ahab died a natural death because of the formulaic statement in 1 Kgs
22.40 that 'Ahab reposed with his fathers'.77 However, since the same for-
mula is used to mark Jehoahaz's death in 2 Kgs 13.9, one wonders whether
this is simply the result of editorial confusion over the various narratives
and traditions, or whether the author-compilers of Kings ever actually
intended the formula to mark a natural death. It is clear that Ahab died in
the same year he fielded a force to fight Shalmaneser III at Qarqar in 853
BCE. That this is the case is confirmed by the fact that Shalmaneser records
receiving tribute from Jehu in 841 BCE. The biblical chronology places 12
years precisely between the end of Ahab's reign and the beginning of
Jehu's reign. This being the case, Ahab's death in 853 BCE may have
occurred on the field of battle against Shalmaneser III at Qarqar or as the
result of an injury suffered at this battle. The memory of this event may
have been fused with the memory of Jehoahaz's death at the hands of Bar
Hadad II.
78. Although it is possible simply to see 'Bar Hadad' as a title and 'Hadad-'idri' as
the personal name of the king of Aram-Damascus, or even that the king had two names
(compare Uzziah/Azariah of Judah), the fact that there is a difference in the name
supports the notion of a misplaced narrative.
79. The exact nuance of the Hebrew word miMn is debatable. The word is derived
from a root meaning 'outside' and could refer to streets, bazaars or distribution centres
for produce and manufactured items. Biran has surmised that the structural complex in
which Fragment A was recycled was actually a ilHin, interpreting this as a marketplace
that stood 'outside' the city walls. See Biran, 'Two Bronze Plaques', pp. 50-52. How-
ever, the nature of this structure in its earliest phase may suggest a military purpose
(p. 45). I have rendered the term as 'outposts' here in an attempt to convey its general
meaning as well as its ambiguity.
286 The Tel Dan Inscription
80. This figure could also refer in the text to the whole population (or male
population) of Israel.
7. Historical Commentary 287
The king of Israel is expressly named only three times in the narrative
(1 Kgs 20.2,13,14). Otherwise, the king of Israel is anonymous. In 1 Kgs
20.13-14 Ahab is named in the encounter with the anonymous prophet. It
is possible that this episode has been inserted into an account of an anony-
mous Israelite king who was besieged in Samaria by 'Ben Hadad'that
is, Bar Hadad. Thus, we may have here the intertwining of two separate
traditions into the one narrative. Indeed, the account of a victory for Ahab
seems quite odd given the agenda of the writer-compilers of Kings to dis-
credit Ahab and the Omride Dynasty. Alternatively, the name of Ahab has
simply been injected into the narrative to provide a specific name for an
anonymous Israelite king.
Two points bear mentioning. First, in 853 BCE Hadad-'idri was the king
of Aram-Damascus who fought with Ahab as an ally on the battlefield of
Qarqar. We also know that Ahab died in that same yearnamely, 853
BCE. If there was a 'Ben Hadad' who was a contemporary of Ahab, then he
must have reigned prior to Hadad-'idri, who was ruling Damascus when
Ahab died. This would make Bar Hadad I, the son of Tabrimmon, the
perfect candidate. However, we then run aground on chronological issues:
how could Bar Hadad I have killed Ahab in battle in 853 BCE if Bar Hadad
I had been succeeded by Hadad-'idri well before 853 BCE? The equation
simply does not hold up. Therefore, the 'Ben Hadad' of the Ahab narra-
tives cannot be Bar Hadad I. Rather, he must be a second Bar Hadad. Yet,
the Tel Dan Inscription shows us that this second Bar Hadad was not
actually a contemporary of Ahab.
We now have a problem arising from the narrative of 2 Kgs 8.7-15,
which twice names Hazael's predecessor as 'Ben Hadad'. We have a num-
ber of possible solutions. First, we could understand 'Ben Hadad' and its
Aramaic equivalent, 'Bar Hadad', as an honorific title for the king of
Aram-Damascus, much like 'Pharaoh' for the king of Egypt. The difficulty
with this, however, is that not every Damascene king was referred to as
'Bar Hadad'. Hazael was certainly not referred to as 'Bar Hadad' and
neither were any of the latter Damascene kings. In fact, epigraphic evi-
dence clearly shows that Hazael was referred to by the title ]K"1Q ('Our
lord'). This title seems to have been used by his son, also, when we see
Adad-Nirari III refer to him as 'Mari'.81 Thus, all the evidence points to
'Bar Hadad' being a proper name as opposed to a title.
The second option is to see 'Ben Hadad' of 2 Kgs 8.7-15 as Hadad-
'idri's successor who reigned for less than four years some time between
845 BCE (the last time Hadad-'idri faced Shalmaneser III in battle) and 841
BCE (the first time Shalmaneser III encountered Hazael). Pitard has sug-
gested this as a possibility.82 Attractive as this solution might be, however,
it goes directly against the Assyrian epigraphic evidence. A summary in-
scription from the reign of Shalmaneser III clearly states that 'Hadad-'idri
died; Hazael, the son of a nobody, seized the throne'.83 This statement
81. See the Tell el-Rimah Stele, line 7; the Nimrud Slab, line 15; and the Saba'a
Stele, line 19.
82. See the lengthy discussion about the entire chronological difficulty associated
with the Bar Hadad question in Pitard, Ancient Damascus, pp. 132-38.
83. For the cuneiform text, see L. Messerschmidt, Keilschrifttexte as Assur: his-
torischen Inhalts (2 vols.; Osnabriick: OttoZeller, 1970), I, p. 30. See also Luckenbill,
Ancient Records, I, 679-83; Kuan, Neo-Assyrian Historical Inscriptions, pp. 53-54.
7. Historical Commentary 289
makes the possibility that a 'Bar Hadad' reigned after Hadad-'idri and
before Hazael extremely slim, if not altogether impossible.
Our third and most plausible option is to replace the name 'Ben Hadad'
in 2 Kgs 8.7-15 with that of'Hadadezer' (i.e. Hadad-'idri). This solution
recognizes the confusion in the biblical text surrounding the 'Ben Hadad'
narratives due to their misplacement, and seeks to amend it in light of the
known geopolitical circumstances of the ninth and eighth centuries BCE.
Hazael, therefore, must be seen as the murderer of Hadad-'idri. Bar Hadad,
Hazael's son, should then be enumerated as the second (and last) Damas-
cene king to bear that name.
some time to head off Jehoahaz's advance, perhaps because he had liter-
ally just been crowned.
Nevertheless, Bar Hadad was able to face Jehoahaz and Joash in battle.
During the battle, Jehoahaz was killed. That Bar Hadad did not actually
capture Jehoahaz and execute him seems to be implied by the fact that
Jehoahaz's son, Jehoash, succeeded to the throne of Israel. This victory
consolidated the Aramaean position in Transjordan and Bar Hadad appears
to have placed a garrison at Lo-Debar, just east of the Jordan River, in
order to check any further Israelite attacks into Gilead. This surmise comes
from the fact that Israel viewed their eventual recapture of Lo-Debar as a
praiseworthy achievement (Amos 6.13).
The memory of this battle at Ramoth-Gilead is preserved in Fragment A
of the Tel Dan Inscription. That it was the first of Bar Hadad's victories is
implied by the fact that he mentions it immediately after praising the mem-
ory of his father, Hazael, who had probably just died before the battle. Bar
Hadad then appears to have followed this victory up with a raid into Israel-
ite territory west of the Jordan. Just how thorough a campaign this was we
cannot fully tell. The memory of an Aramaean siege against Samaria in 1
Kings 20.1-22 and 2 Kings 6-7 probably stems from this time. Whatever
the case, Bar Hadad does boast of some raiding campaign into 'their land'
(Line A10)that is, the land of Israel and Bayt-Dawid. It seems a logical
strategic move if Bar Hadad did indeed lay siege to the Israelite capital at
Samaria.84
If we are permitted to see the Aramaean siege of Samaria as occurring at
this time, then we may also have a reason for why Samaria did not actually
fall. From the Eponym Chronicle we know that on the day of the Assyrian
New Year Festival in the month of Nisan (April) of the year 796 BCE, the
Assyrian king, Adad-Nirari III, was located with his army in the Lebanese
Beqa' at Mansuate. The biblical text also informs us that a 'saviour'
rescued Israel from the Aramaean onslaught (2 Kgs 13.5). This 'saviour'
was undoubtedly Adad-Nirari himself. If we are able to tie all these threads
together, then the Aramaeans may have lifted the siege of Samaria because
of the Assyrian threat in the north. We notice that a similar sentiment is
expressed as the reason for why the Aramaean siege of Samaria was lifted
in one of the Elisha narrativesnamely, 2 Kgs 7.6.
84. Compare Sasson, who first mentioned the possibility of a siege against Samaria
being recounted in the Tel Dan Inscription, though he based it partly on an unlikely
restoration stemming from a misreading of Line Al. See Sasson, 'The Old Aramaic
Inscription from Tell Dan', p. 15.
7'. Historical Commentary 291
The probability that Bar Hadad laid siege to Samaria is strong, not only
because of the presumed historical-strategic considerations, but also
because of the content of the Tel Dan Inscription itself. We note that Line
A13 does actually mention a siege. Unfortunately, the besieged city's
name is no longer extant. It is my contention, however, that since a siege
of Samaria would certainly have rated a mention and was probably Bar
Hadad's primary goal, and since the memory of an Aramaean siege of
Samaria by 'Ben Hadad' is preserved in the biblical text, we can recon-
struct the missing name of the city in Line A13 as |~1D^ ('Samaria') with a
fair degree of certainty.
That the siege was ultimately a failure, or at least did not lead to any
wide destruction to Samaria, is evidenced by Samaria's archaeological
record, which shows no conflagration from this era. This is backed up by
textual evidence. The Tell er-Rimah Stele, which records that Jehoash of
Israel paid tribute to Adad-Nirari III, refers to the Israelite king as ' Joash
of Samaria' (Akkadian: lasu KmSamerina). The memory of an unsuccess-
ful siege of Samaria by 'Ben Hadad' is also quite strong in the biblical text
of Kings. The siege's failure does not preclude it being referred to in the
Tel Dan Inscription, though, for in the Tel Dan Inscription we are dealing
with royal propaganda. We may compare it with Sennacherib's claim to
have caged Hezekiah up like a bird in Jerusalem, yet without reference to
Jerusalem's wall being breached or the gates penetrated. An Aramaean
siege of Samaria at this time does make sense when we consider the possi-
bility that Jehoash of Israel probably appealed to Adad-Nirari III at this
time to come to his aid, indicating that Jehoash found himself in dire
straits. The Assyrian king's willingness to check Bar Hadad, then, was the
reason for Bar Hadad's retreat from Samaria,85 as well as the reason why
Jehoash is listed as having paid tribute to Adad-Nirari III.
Meanwhile, Joash of Bayt-Dawid, who seems to have survived the ear-
lier battle of Ramoth-Gilead, was assassinated in his capital city during a
palace coup (2 Kgs 12.20-21). This would have occurred approximately
two years after the battle of Ramoth-Gilead, probably in late 797 BCE o
early 796 BCEthat is, just before or around the same time Adad-Nirari III
was campaigning in southern Syria. Joash's son, Amaziah, was installed as
king of Bayt-Dawid in Joash's stead. It is Amaziah's name that appears
partially preserved in Line B8 of the Tel Dan Inscription.
The context of this reference appears to have been another battle.
Whether this occurred before or after Bar Hadad paid a hefty tribute to
and two top stones) must certainly have been an oversupply for whoever
may have lived in this structure. It seems far more plausible to suggest that
this was an industrial installation, perhaps that of a baker.
This suggestion gains more weight when we consider the finds in other
parts of the building, such as an oven and numerous storage jars, kraters
and jugs.91 One room, which Yadin identified as a bedroom (Room 44a),
even contained a pithos, two storage jars and two cooking pots, as well as
some iron tools. This hardly befits the description of a bedroom. Interest-
ingly, an intricate ivory cosmetic spoon was also found at the doorway.92
Yadin remarks that this dainty 'is one of the few of its kind to be found,
not in palaces, but in a private house'.93 However, we suspect that Build-
ing 14a was not a house at all, but a bakery, and that the cosmetic spoon
belonged to someone who probably did not live in Building 14a but who
may have frequented it.
Makbirram, therefore, should probably be seen as a client of this bakery
rather than a resident. That he had pottery custom-made for him and that
he may have been supplied in bulk from this bakery attests to the fact that
Makbirram was a man of some standing and wealth. Interestingly, Build-
ing 14a is located adjacent to Building 2a, which Yadin describes as 'the
most beautifully planned and preserved building among the Israelite struc-
tures of Hazor'.94 Did this serve as Makbirram's residence? Whatever the
case, Makbirram's prominence is secure. Whether he was a governor, petty
king, general or merchant is unknown. Whether he was the personage
mentioned in Line B7 of the Tel Dan Inscription is also beyond absolute
proof. From the stratigraphy of Hazor, however, we know that Makbirram
was a contemporary of Bar Hadad II.
Bar Hadad seems to imply from Fragment B of the Tel Dan Inscription
that he was victorious in the battle alluded to in Line B2. If this battle had
been against Adad-Nirari III and his army in 796 BCE, as I have suggested,
then we must question Bar Hadad's claim here. From the Tell er-Rimah
Stele, it is clear that Bar Hadad (also known as 'Mari') was the loser of
any armed encounter with the Assyrian king. The thoroughly documented
tribute that he brought to Adad-Nirari III testifies to this. If, therefore, the
stele of the Tel Dan Inscription had been erected in response to an encoun-
ter with Adad-Nirari III, it may have been an act of damage control on the
part of Bar Hadad. That is, since Bar Hadad was so weakened by the
humbling encounter with Adad-Nirari, he may have thought it expedient to
reassert his strength in his various territories with an inscription describing
his own prowess and authority in parallel with that of his illustrious father,
Hazael. Thus, the Tel Dan Inscription may represent Bar Hadad's propa-
ganda in the face of his waning power and a last-ditch attempt to assert his
sovereignty over Galilee and Transjordan.
Alternatively, if Bar Hadad actually was successful in this second battle,
it does not appear to have been fought against Adad-Nirari. In this case, it
may have been fought against a small coalition of forces trying to over-
throw his suzerainty. Yet, since this battle most likely took place after his
submission to Adad-Nirari, one suspects that Bar Hadad could not have
held out against such a coalition. Indeed, Jehoash's three victories over
Bar Hadad attest to the fact that the Damascene king was left almost pow-
erless after his encounter with Adad-Nirari. Therefore, once again, we may
surmise that Bar Hadad erected the Tel Dan Inscription in the vain attempt
to stamp his authority over the regions that were rapidly slipping from his
grasp. It may have even been a response to one of Jehoash's victories over
him.
Whatever the case, it appears that the Tel Dan Inscription represents
deliberate propaganda by Bar Hadad to promote his suzerainty over former
Israelite territories in the face of his waning influence over these territo-
ries. It was presumably aimed at Dan's leaders, as well as the leaders of
other major centres in Transjordan, Galilee, and in Israel proper, in an
attempt to dissuade rebellion against Bar Hadad, as well as any association
with Jehoash of Israel. Since Dan was in a geographical (and perhaps
political) 'no-man's land' between Israel and Aram-Damascus, its gates
were the most suitable place to erect the stele. Dan may also have been the
only external territory that Bar Hadad still held at the time the Tel Dan
Inscription was composed. In any event, the Tel Dan Inscription is a cover
for Bar Hadad's rapid demise. The subsequent revival at Dan and other key
sites like Hazor demonstrates Jehoash's success at expelling Bar Hadad
from both Transjordan and the Galilee. The lack of evidence for any con-
flagration at Dan suggests that it had been transferred to Israelite control
without any siege or major hostility at Dan itself. This would corroborate
well with the information about Jehoash's three victories at Apheq, Lo-
Debar and Qarnayim, in that these victories were enough to expel Bar
Hadad and the Aramaean forces from all of Galilee and Transjordan.
7. Historical Commentary 295
Timeline
The sequence of events surrounding the production and destruction of the
Tel Dan Inscription may be reconstructed as follows:
the word llimi in the Tel Dan Inscription because even the previous data
and conclusions are so hotly debated. The value of the biblical texts for
historical reconstruction becomes a crucial issue, for it is primarily with
the biblical narratives of David that we must interact as we investigate the
derivation of the element 111.
Etymologically and contextually, the Aramaic toponym HlfTD (Bayt-
Dawid) is identical to the Hebrew toponym 111 TU ('Ir Dawid, or 'City of
David'). The biblical texts make the statement that Jerusalem was captured
by David and renamed 111 TU ('City of David') after himself (2 Sam. 5.7,
9; 1 Chron. 11.7). However, many, such as Davies,97 Lemche98 and Thomp-
son,99 argue that for methodological purposes, the biblical narratives on
David, which must be dated to the Persian era at the earliest, cannot be used
for reconstructing the history of Syria-Palestine in the early tenth century
BCE. Rather, it must be the role of archaeology to elucidate the history of
Syria-Palestine, especially in the early Iron Age, which has yielded next to
nothing of a literary nature. If this approach is taken, then the biblical
account of how the toponym "111 TI? ('City of David'), and therefore its
Aramaic equivalent 11 Iff 3 (Bayt-Dawid) was derived loses its value as a
historical source. The temporal gulf between the early Iron Age and the
setting down of the traditions in writing makes the biblical record an
unknown quantity at best and a pure fabrication at worst.
Thompson, however, does analyze the biblical texts. He does this in
order to ascertain contextually how the name 'David' (111) was under-
stood by the biblical writers themselves, late though they may have been,
and to see if this gives any clue as to the derivation of the name at an ear-
lier time.100 According to Thompson, the issue is one of priority: whether
the name Bayt-Dawid (llinTl) was derived from the name of a personage,
'David', or whether the name 'David' was derived from the word Bayt-
Dawid (HIST 3). If the latter were the case, then the name 'David' must be
derived from an epithet of Yahweh. It is at this point that Thompson turns
to the biblical texts to support his suggestion that this latter scenario was
indeed the case.
translation in keeping with the meaning of the Semitic root Vlll, but it
makes good contextual sense. A deity, therefore, could legitimately be
given the title "111. However, biblical texts demonstrate that it is also used
as a human appellation, most notably in the personage of David (~n~[ or
T11). In this regard, the biblical name 'David' can be seen as a title that
takes on the significance of a personal name, similar to an example cited
by Thompsonnamely, 'Bar Hadad' derived from 'Son of Hadad'.102
Yet, Thompson argues that the biblical character David is the personifi-
cation of Yahweh's patronage of Jerusalem's kings. Thompson bases this
assertion on the patron language within the biblical narratives themselves.
However, this suggestion does not appreciate the complexity of patron
language in antiquity. The patron language surrounding the title or per-
sonal name TO suggests that it arose not from the notion of Yahweh's
patronage of Jerusalem's kings, but from these kings' patronage of other
people, especially of the people of Jerusalem and Judah. Indeed, it is in
this capacity as royal patrons that the House of David is often addressed in
biblical literature. Within the label TH mi ('House of David') the ele-
ment TH refers to the patriarch of an aristocratic family landed in Jerusa-
lem (as well as Bethlehem and Hebron) who was acknowledged with the
title "pft ('king'). In other words, the label TO JTD ('House of David')
does not see the kings of Jerusalem as the subordinates in a relationship
with their deity, but as the superiors in a relationship with their 'subjects'.
The House of David is only seen to stand for Yahweh through the use of
metonymy.
Certain key texts provide the evidence.
2 Samuel 7.5-16. This text relates the divine oracle given to David about
the establishment of David's house. The language is very clearly that of
Yahweh's patronage of David and his descendants. Although the term JTm
"TH is never used, it is implied numerous times. Of particular importance is
the consistent parallelism between the terms PPD ('house') and rD^QQ
('kingdom'). This indicates that the label 111 mi did not refer primarily to
Yahweh's patronage of Jerusalem's kings, but to the kings' patronage of
the town of Jerusalem. In distinction, Yahweh's divine patronage of Jerusa-
lem's kings is described with different patronage terminologynamely, DK
('father') and p ('son'). That is, the relationship between Yahweh and Jeru-
salem's kings was not one of extended family (m, 'house') which saw
1 Kings 12.16-33. This famous text deals with the response of the Israelite
tribes to Rehoboam's harsh demands, as well as Jeroboam's subsequent
construction of shrines at Bethel and Dan. The Israelite tribes are said to
rebel against the House of David (1 Kgs 12.19). If this actually referred to
Yahweh, then the rhetorical question on the Israelites' lips, 'What share do
we have in David?' (1 Kgs 12.16), makes little contextual sense. Indeed,
Jeroboam's program of building new Yahwistic shrines would seem to
defeat the purpose of this complaint which leads to secession and his
subsequent desire to prevent the kingdom returning to the House of David
(1 Kgs 12.26). Rather, the rebellion of Israel is against the king of Jeru-
salem, Rehoboam, and his patronage of the tribes. The only way we can
see the House of David as referring to Yahweh in this text is through
metonymy, and it certainly seems that the text intends the reader to make
that metonymic association. That is, in rebelling against the House of
David, the Israelite tribes rebel against Yahweh by effectively questioning
their own share in Yahweh. Yet, they rebel not just against Yahweh as
patron of Jerusalem's kings, but as divine patron of all the Israelite tribes.
However, the prime referent of the name 11 "T is the human king of
Jerusalem, not the divine one. Only through metonymic association can
there be a connection between Yahweh and the House of David. Should
this metonymic association fail, the term "TH ITU ('House of David') is
still meaningful as a reference to the kings of Jerusalem, exemplified speci-
fically here by Rehoboam.
describes in this utterance as his faithful servant. This cannot refer to the
man who is king in Jerusalem at this particular point in the narrative (Reho-
boam) because he is earlier portrayed as a fool in order to accomplish
Yahweh's turn of events (1 Kgs 12.15). Rather, 111 must refer back to the
eponymous ancestor of Jerusalem's kingnamely, David ben-Jesse.
Ophel Ridge between the Qidron and Tyropoean Valleys, the remains of
the site resemble a small fort rather than an actual town or village. Excava-
tions have yielded meagre remains, both in terms of structures and wares.
Jerusalem at this time appears to have been a humble, even backward,
family estate. As such, what seems to have made Jerusalem famous, or
rather notorious, was not the physical town itself, if we can even call it a
town during this period. Rather, it was its few inhabitants that gave Jeru-
salem any sort of relevance on the political stage. Evidence suggests that
Jerusalem, known at this time as Til "I1117 (or TUfTD in Aramaic), was
probably the home of an aristocratic family with a notorious reputation or
pasta kind of self-important 'mafia' clan.
In comparison, the biblical record presents a very similar picture. Tradi-
tional notions down through the ages and into our time have built up fanci-
ful pictures of Jerusalem which, on closer inspection, do not actually reflect
the biblical picture itself. 1 Samuel 5.6-9 portrays David as living in his
newly conquered compound and renaming the compound from jTiJ ('Zion')
to TH TI7 ('City of David'). This particular pericope corroborates quite
well with the archaeological record. It is clear from archaeology that in the
tenth century BCE, Jerusalem was only a fortified compound with a small
populationnot a bustling metropolis. The Tel Dan Inscription fills this
picture out more by showing that at c. 800 BCE, Jerusalem was known by
another name, TnrP3 (or TR "IT in Hebrew).
Furthermore, it is noteworthy that in the book of Kings, the usage of the
name 111 T JJ ('City of David') ceases with the reigns of Ahaz and his son
Hezekiah. The name is most often referred to in the biblical texts as the
burial site of Jerusalem's kings. Ahaz is the final king explicitly said to
have been buried in TH TU, the 'City of David' (2 Kgs 16.20), though the
Chronicler may give slightly different information when he puts Ahaz's
burial place 'in the city, in Jerusalem' (2 Chron. 28.27).104 Hezekiah's
burial is glossed over without details in 2 Kgs 20.21, but the Chronicler
tells us that he was buried 'in the upper of the tombs of David's sons'
104. Exactly what the Chronicler meant by this phrase is difficult to ascertain. It
could be inferred that Ahaz was not buried in the royal compound, ~IH TU, but was
rather buried in another part of the burgeoning city of Jerusalem. This would contradict
with the information given in 2 Kings. Alternatively, the Chronicler could be stating
that Ahaz was buried 'in the city', referring to the royal compound, "Til TU, which he
further describes as being situated 'within Jerusalem'. In that case, the Chronicler
would agree with the information given in 2 Kings.
306 The Tel Dan Inscription
(2 Chron. 32.33). Presumably, this means Hezekiah was also buried in the
royal compound known as TO "IT ('City of David').
During the reigns of both Ahaz and Hezekiah (c. 732687 BCE), Jeru-
salem was expanding beyond the fortified compound and the entire Ophel
Ridge. Jerusalem reaped the economic benefits of its expanding sover-
eignty during the eighth century BCE and the site expanded from a simple
family compound to the capital city of a bustling small state. It was during
Hezekiah's reign that the city was besieged by Sennacherib (701 BCE) who
recalls the name of the city as Ursalimu ('Jerusalem').105 It was during this
time that the new suburbs of Jerusalem were walled around and the water
system expanded. From this we can deduce that the name TH "IT ('City
of David') referred in the strictest sense to the fortified compound of
Jerusalem's kings, but was used in a wider sense as a reference for all of
Jerusalem, especially at a time when all of Jerusalem hardly exceeded the
compound itself. The name 'Jerusalem' again came into vogue by the time
of Hezekiah (late eighth century BCE) when the city had expanded into a
true urban entity beyond the royal compound. This means that the name
'Jerusalem' was preserved throughout this period as the general name of
the location of the fortified compound. Since Jerusalem can hardly be
called a city before the eighth century BCE, it was better known during this
period by the name of the fortified compound that was its main feature.
Furthermore, the progression of kings in the biblical record appears
accurate when we bring the archaeological record to bear on it. This means
that there must be a very real connection between the writer(s) of Samuel-
Kings, as well as the Chronicler, and the kings themselves. If many of
Jerusalem's kings were entombed within the original royal compound
the 'City of David'then the connection becomes clearer: the royal
compound itself, down to the destruction of Jerusalem in 586 BCE, must
have been a veritable historical capsule containing the remains of numer-
ous generations of Jerusalem's rulers. Although this probably does not
completely bridge the gap between Jerusalem's kings and the biblical
writers, it certainly closes it considerably.
Thompson's assessment of the connection between referee and referent
is, however, far more conservative:
Continuity in historical records and historical knowledge in ancient
Palestine has been broken, and it has been broken repeatedly. Therefore the
historical referent of our earliest usage within the tradition to a Davidic/
105. ANET,pp.2S7-l++
7. Historical Commentary 307
Solomonic empire may go back only as early as our sources allow us to.
Their historical context also belongs to the Hellenistic period, when our
texts' referents to this empire first unequivocally took a form comparable to
what they have now. However, this referent could far more easily be to the
Hasmoneans than to someone living in the remote and nearly wholly
unknown tenth century...106
BCE king of Israel. Of course, the character of Omri in the biblical text
serves a different function to the Omri in the Mesha Stele, but the referent
of both is undoubtedly the same. The different functions of Omri within
these two contexts do not derive from two different referentsone in ninth
century BCE Moab and the other in second century BCE Judaea. Rather, th
different functions and characteristics of each derive from the two differ-
ent milieus that lay behind the writing of each text.
When it comes to the biblical figure of David, we can be confident that
the referent is not an entity in the author's own era, nor a figure purely per-
sonifying Yahwistic patronage for the kings of Jerusalem. This is espe-
cially so given the accuracy of the catalogue of rulers for Judah and Israel,
not to mention the catalogues for other neighbouring states. Rather, we can
be confident that the referent is a historical personage and one who lived in
the early Iron Age. We may need to revise our previous understanding of
the circumstances surrounding this David; he may well be the warrior
chieftain whom Jamieson-Drake surmises roamed the Judaean hills in the
tenth century BCE and who gave rise to the Davidic traditions of the bib
lical texts,107 rather than the imperial persona at the head of a great civili-
zation as we are often led to believe. Since Jerusalem was evidently a small
feudal estate in the late ninth to early eighth century BCE, this is a reason-
able scenario.
Having said this, however, we must be very cautious with the Tel Dan
Inscription. Its author, Bar Hadad II, was simply not interested in telling us
about David. Rather, he was interested in propagating his version of events
in his own lifetime c. 800 BCE. Also, the lexeme "inn''!} (Bayt-Dawid)
refers not to a dynasty, but to the settlement in Jerusalem at that same time.
As such, David is simply not in the inscription's scope. The inscription
does, however, bring us a step closer to David by filling in more of the
picture of Syria-Palestine during the Iron Age. The inscription presents us
with a location incorporating the word TH in its name, and Jerusalem is
the logical identity of that location. To say that David is the original refer-
ent of that word, as the biblical literature purports, is certainly a logical
conclusion to make, but it is one which cannot be proven beyond a doubt
as the actual permutation that occurred. It is certainly plausible, though.
To put it another way, the Tel Dan Inscription neither confirms nor denies
the biblical assertion that a certain David lent his name to a fortress in Jeru-
salem. This, however, is because that naming event is outside the inscrip-
tion's scope, the inscription being interested in other matters. Given the
evidence, however, it certainly looks as though it was the case. As such,
we cannot say that we have pinned David down outside the pages of the
Bible. We may well, so to speak, have found a footprint, even a fresh one,
but he himself still eludes us. We are, however, hot on his heels and our
confidence in finding him has greatly increased. The Tel Dan Inscription
does not give us proof of an historical David, but it may certainly be
admitted as evidence.
Cultic Implications
The Tel Dan Inscription alerts us to certain cultic developments that were
occurring in Syria-Palestine at the time the inscription was producedthat
is, in the first years of the eighth century BCE. These developments have to
do with the nature of the cult of El, as well as that ofmassebot,++++++
stones.
In the Tel Dan Inscription, we find a reference to the deity El-Baytel in
Line A4. This deity appears to be connected with the later deity, Bethel.
The name of Bethel appears in Esarhaddon's treaty with Baal, king of
Tyre (IV 6; c. 676 BCE), and in his Succession Treaty (line 467; c. 672
BCE).108 It also appears as a theophoric element in many Aramaic personal
names throughout the Near East from the sixth century BCE onwards.
Bethel is, in all likelihood, an evolution of the more archaic deity, El. This
is supported by the fact that ^K ('El') largely disappears as a theophoric
element in Aramaic personal names after the seventh century BCE,
replaced instead by ^n-3 or ^Nnn ('Baytel'/'Bethel').109
Since the Tel Dan Inscription may be dated to c. 796 BCE, it is
particularly useful in tracing the development of this deity from the more
archaic 'El' to subsequent 'Bethel'. The name [^KH^Il^N ('El-Baytel') in
Line A4 should be seen as a stage in the deity's evolutionary development.
The fact that Esarhaddon entreats Bethel in his treaty with Baal, king of
Tyre, implies that by the early seventh century BCE, the cult of Bethel was
well established in the region of Phoenicia. If Jer. 48.13 is historically reli-
able, then we may presume the presence of the Bethel cult in Israel by the
time of the Assyrian destruction of the state in 722 BCE. The temple and
cult of Bethel among the Judaean community at Elephantine and Syene
also attests to the cult's existence in Palestine during and probably before
the sixth century BCE.
There is also good reason to suggest a connection between this deity and
sacred Bethel-stones (or 'Baetyls') which were venerated at cultic shrines.
The term 'Bethel' for such sacred stones is well known from both Aramaic
and later Greek texts where it appears as an element in numerous com-
pound names of deities. These include 'Anat-Bethel,110 Ishum-Bethel111
and Herem-Bethel112 in the Aramaic papyri of Elephantine, and Zeus
Betylos,113 Zeus Baitylos114 and Symbetylos115 in later Greek inscriptions
from Syria and Mesopotamia. Baitylos is also explicitly named as a
'stone-god'.116 The fact that 'Anat and Ishum appear in compound divine
names from Elephantine confirms that we are dealing with sacred Bethel-
stones, for these are also known as independent deities in Syria and Pales-
tine without compounds.117 Thus, the compound divine names may be
understood as appositional constructions. So, 'Anat-Bethel means 'The
Bethel-stone, 'Anat'.118 In regard to the divine name 'El-Baytel' in the Tel
Dan Inscription, we may understand this name in the same way, meaning
'The Bethel-stone, El'.
Sacred Bethel-stones are well attested in the Levant and in the Phoeni-
cian colonies. In all probability, the massebot uncovered at numerous
Levantine sites, including Tel Dan, were venerated as sacred objects. The
term^Kmi ('house of god' or 'divine house') may have been a reference
to the entire shrine in which a massebah stood. These massebot, or Bethel-
stones, seem to have been venerated as potent stones that were essentially
aniconic representations of a deity. Thus, the presence of more than one
massebah at a shrine is indicative of the veneration of more than one deity
at that shrine. Their significance may be likened to that of icons in Eastern
Orthodox religion.119
The Jacob narratives are particularly instructive here as to the nature and
function of Bethel-stones or massebot. After his nocturnal vision of Yah-
weh at Luz, Jacob sets up the stone on which his head rested as a masse-
bah. He then takes a vow concerning this massebah.m The implication of
this is that the massebah acts as a witness to his vow. Throughout the Bible,
stones (and altars) are set up as witnesses to vows, treaties and memorable
events and treated as shrines.121 Other texts, such as Gen. 31.13,122 Jer. 2.27,
Dan. 2.34-35122 and Zech. 3.9 also demonstrate how these sacred stones
were considered animate objects.124
money than Yahu. Ishum-Bethel, however, receives much less than both the other
deities. See also Cowley, Aramaic Papyri, p. xviii; K. Van der Toorn, 'Anat-Yahu,
Some Other Deities, and the Jews of Elephantine', Numen 39 (1992), pp. 80-101.
119. This is not to suggest an equivalence between sacred Bethel-stones and
Orthodox icons. Rather, icons are mentioned here as a latter day parallel to help
explain the significance of sacred Bethel-stones for the ancients of the Levant.
120. Gen. 28.18-22. It is ironic that Jacob then names that place 'Bethel'. It is
questionable, however, whether this appellation has anything specifically to do with the
massebah or sacred Bethel-stone which Jacob erects. The name is connected with
Jacob's vision rather than the massebah. However, the connection may have been
misconstrued, either accidentally or deliberately, in the aetiology.
121. E.g. Gen. 31.44-54; Josh. 4.1-9; 22.21-29; 24.26-27; Judg. 13.19; 1 Sam. 4.1;
6.14-15; 7.12; 1 Kgs 1.9; Isa. 19.19-20.
122. This verse is commonly mistranslated as a construct expression, 'I am the deity
of Bethel'. The syntax, however, suggests that it should properly be rendered
appositionally, 'I am the deity Bethel'.
123. This verse may carry overtones of Greek mythology in which either Kronos or
Ouranos ('heaven') endows certain stones ((3aiTuAta) with life. These particular
stones are considered to be of divine origin, perhaps meteoric stones; see M. Ugolini,
'II dio (di) pietra', Sandalion 4 (1981), pp. 7-29. Damascius (Vita hid. 94, 203) uses
312 The Tel Dan Inscription
the same term (^aiTuXia) for certain stones which fell from the sky in the Lebanon.
See also Van der Toorn, Becking and Van der Horst (eds.), Dictionary of Deities and
Demons, pp. 299-300.
124. Isa. 8.14 and Hab. 2.11 may also encapsulate the idea of animated stones.
125. Biran, Biblical Dan, pp. 238-45.
126. Compare Gen. 28.18; 31.46; 35.14; Deut. 4.28; 28.36,64; 29.17; Judg. 9.5,18;
13.19; 1 Kgs 1.9; Jer. 3.9; Ezek. 20.32.
127. This attribute of El is clearly seen in Ugaritic literature. See M. Dietrich, O.
Loretz and J. Sanmartin, Die keilalphabetischen Texte aus Ugarit (Neukirchen-Vluyn:
Neukirchener Verlag, 1976), KTU 1.3.V.30; KTUlA.ivAl, v.3-4; KTU 1.16.iv.l-2.
128. Biran and Naveh, 'The Tel Dan Inscription', pp. 1-3, esp. Figs. 1 and 2.
7. Historical Commentary 313
(^"18), or even that the entire platform was considered a sacred hearth.129
Since Fragment B1 was found in debris just 2 m south of this construction,
the stele of the Tel Dan Inscription may have originally stood next to it.
This would provide a tangible connection between the content of the
inscription and its archaeological context.
Also interesting in explaining the function of sacred stones are the texts
of Gen. 35.19-20 and 2 Sam. 18.18. In Gen. 35.19-20 we read that after
Rachel's death, Jacob set up a massebah on or next to her grave. Thus,
sacred stones were also considered to perpetuate the memory of a deceased
person. This is also clear from the text of 2 Sam. 18.18, where we read that
Absalom, during his lifetime, had set up a massebah in order to preserve
the memory of his name because he had no son. Interestingly, he is said to
have named the massebah after himself. This is particularly noteworthy for
the Tel Dan Inscription. In wishing that his father be remembered 'at every
ancient hearth on [sacred] ground of El-Baytel', the author, Bar Hadad, can
be understood as expressing the desire to see a sacred Bethel-stone erected
at every such hearth in order to perpetuate his father's memory. It might
also indicate that the deceased king was considered to have joined or
merged with the deity, El-Baytel, in his death. We see this belief at Ugarit
where deified kings 'joined' El in the afterlife. 13 It seems plausible that this
belief was carried on considering the connection between El and El-Baytel.
Alternatively, we may understand Bar Hadad's wish here as simply to
have his father, Hazael, remembered at every ancient hearth of the deity
El-Baytel. In wishing this, Bar Hadad may have desired that his father be
remembered at all proceedings in the city gate, presumably as a wise and
just king, so that his memory could be perpetuated by all those at the city
gate.
The Tel Dan Inscription affords us a glimpse into the significance of not
only the cultic platform at Dan, but also those found at the gates of other
Iron II cities. Of particular note is the cultic platform found in Area A at
Bethsaida, from Stratum V (ninth-eighth centuries BCE).131 This cultic
platform measures 2.31 x 1.53 m at the base. Two steps, however, lead up
to a raised square platform that is 1.53x1.53m in area. The platform was
129. Biran and Naveh call the entire platform a bamah (HQ3). However, bamot were
most likely larger sacred complexes such as that in the northwest corner of Tel Dan.
130. J.C. de Moor, The Rise ofYahwism (Leuven: Leuven University Press, 1990),
p. 242. Compare also the idea of deceased kings dining with Hadad in Panammu I,
lines 17-18.
131. R. Arav, 'Bethsaida 1996-1998', IEJ49 (1999), pp. 128-36 (132-34).
314 The Tel Dan Inscription
located at the base of the northern gate tower of Bethsaida. In the centre of
the raised platform, excavators found a rectangular basalt basin in which
two incense burner cups were found. This find probably indicates that the
basin was a sacred hearth (^K~IK). It is also evident from a depression at
the far end of the platform that the basin stood before a massebah. A deco-
rated iconic massebah found nearby had been deliberately knocked down
and broken into five pieces. The relief on this iconic massebah has been
interpreted as an Aramaean manifestation of the moon god.132 However, it
is debatable whether this massebah was the one originally located on the
platform.133 Whatever the case, it is evident that these massebotwere asso-
ciated with legal proceedings at the city gate, as well as the memory of
prominent deceased persons.
The fact that the name of'El' is largely replaced by 'Baytel' or 'Bethel'
from the seventh century BCE onwards suggests that El came to be known
as 'El-Baytel', meaning 'The Bethel-stone, El', prior to the seventh cen-
tury. This is because El would have been commonly represented by a
sacred Bethel-stone at the city gate, which acted as a witness to legal
proceedings there. The cult of El, as represented in specific temple com-
plexes, seems to have waned in the Iron Age because of the dominance of
the Baal cult throughout the Levant. However, the attributes of El allowed
his cult to be perpetuated on a lesser scale in a more nominal role as the
proverbially wise and just god who witnessed legal proceedings from the
shrine at the city gate. As such, El came to represent wisdom and justice in
his form as El-Baytel and Bethel.
This personification of El as a sacred Bethel-stone, presumably the chief
Bethel-stone at the city gate, probably led to the subsequent apocopation
of'El-Baytel' to simply 'Baytel'. Thus, from the seventh century BCE on-
wards, reference to the deity 'Baytel' or 'Bethel' without any prefixed
compound element should be understood as a reference to the deity El in
132. Arav, 'Bethsaida 1996-1998',p. 134; M. Bernett and O. Keel, Mond, Stier und
Kult am Stadttor: Die Stele von Betsaida [et-Tell] Unter Mitarbeit von Stefan Mtinger
(Gottingen: Vandenhoeck & Ruprecht, 1998).
133. R. Arav, 'Bethsaida 1996-1998', p. 134, Fig. 5. Arav says that this massebah,
found at the bottom of the platform's stairs, was the one before which the incense basin
stood. Oddly, however, the length and width of the massebah'?, base (59 x 31 cm) is
too large to fit into the depression on the platform (53 x 35 cm). The depression on the
platform also measures 10 cm in depth. At this depth, the bottom portion of iconic
relief on the massebah would have been obscured from view. If this massebah was
indeed the one originally located on the platform, then we must surmise that it was
placed on some kind of plinth that was fitted into the depression.
7. Historical Commentary 315
his evolved state. Other deities whose names are compounded with bwru
('Bethel') may, however, be understood as distinct from the deity Bethel,
their names being appositional constructions as mentioned previously.
Thus, Esarhaddon's call upon Bethel and' Anat-Bethel in his treaty with
Baal, king of Tyre (IV 6), and in his Succession Treaty (line 467), should
be understood as a call to El and ' Anat in their representations as Bethel-
stones and witnesses to the treaty, to deliver the defaulting parties 'to the
paws of a man-eating lion'.134 From Elephantine, the letter of Yedonyah
and his colleagues to Bigwai, governor of Judaea, informs us that the tem-
ple at Elephantine had stone pillars.135 These are specifically mentioned
first among the objects destroyed by the forces of Waidrang and Nephayan
in their pillage of the temple. One wonders whether 'the pillars of stone'
(NHK 'T KmaU) 136 in the temple were sacred Bethel-stones representing
the five deities mentioned specifically as revered by the Judaean commu-
nity in Elephantine.137
In any case, the Tel Dan Inscription provides us with a key insight into
the evolutionary development of the deity El. It also affords us a valuable
glimpse into the cultic significance ofmassebot, especially those found at
various Iron Age gate complexes.
134. See Parpola and Watanabe, Neo-Assyrian Treaties and Loyalty Oaths, pp. 27,49.
135. See Cowley, Aramaic Papyri, 30.9-10.
136. Note the definite article accompanying the word 'pillars'.
137. These five deities were Yahu, 'Anat-Bethel (or 'Anat-Yahu), Ishum-Bethel,
Herem-Bethel and Bethel (that is, the evolution of El-Baytel). Although the entire
temple complex seems to have been dedicated to Yahu, god (or lord) of the Heavens,
the five gates of this temple would seem to correspond with the five deities of the
community. As such, we expect that each of these deities were represented by an
object. The compound element T>KrPD ('Bethel') in the names of most of the deities
suggests that they were represented by sacred Bethel-stones. See also Cowley, Aramaic
Papyri, p. xviii.
Chapter 8
CONCLUDING REMARKS
The Tel Dan Inscription presents us with a very valuable text to shed light
on the history of Syria-Palestine. Perhaps one of its greatest contributions
to the field of ancient Near Eastern studies is in highlighting the dem+ise o
Aram-Damascus as the dominant power of the Levant and the subsequent
rise of Israel. One has to view the inscription as a work of propaganda for
the regime of Bar Hadad II. Unable to maintain control of the southern
trade routes through Palestine and Transjordan, Bar Hadad resorted to a
propaganda campaign that highlighted his prowess in previous military
campaigns as well as drawing a comparison with his enormously success-
ful father, Hazael. Bar Hadad's claims must be viewed in the light of the
geopolitical developments which both Assyrian and biblical texts bring to
light. As such, we gain a picture of Aram-Damascus' diminishing power.
The inscription also further defines our picture of ancient Jerusalem. In
c. 800 BCE Jerusalem is seen to have been little more than a fortifie
compound belonging to an ancient family. It did, however, form its own
political unit and was recognized as a distinct unit by Bar Hadad. We can-
not talk of a regional Judaean state at this time. At best, Jerusalem was a
principality with very limited sovereignty. Yet in less than a century after
the composition of the Tel Dan Inscription, this small estate had gained
suzerainty over the entire region of Judah and so evolved into a fully
fledged state entity.
Epigraphically, the Tel Dan Inscription is one of the best-preserved
pieces from antiquity and also one of the neatest, even if it is not one -of
the most complete artefacts. The discovery of Fragments Bl and B2
almost a year after the discovery of Fragment A raises hopes that more
fragments can be found. However, as the seasons of excavation continue at
Tel Dan, this becomes a less likely prospect.
With regards to the biblical texts, the Tel Dan Inscription demonstrates
that there are definite historical kernels in the Bible that cannot readily be
8. Concluding Remarks 317
in the Levant would cause such excitement among scholars and students.
However, with future finds we must take care not to fan our excitement
into either hysteria or parochialism. The Tel Dan Inscription produced one
of the biggest divides in the scholarly community. In the attempt to defend
certain perspectives, the integrity of the investigative process was too
often compromised, as was decorum sometimes foregone.
The primary lessons that we must learn are with regard to the investiga-
tive process itself. First, although resources are often frustratingly limited,
scholars must not be tempted to make conclusions over inscriptional arte-
facts based on photographic or hand-drawn images while the actual inscrip-
tional remains are still extant. As was the case with the Mesha Inscription,
there are occasions when scholars cannot rely on anything but manufac-
tured images of the inscription because of destruction to the artefact itself.
However, in the case of the Tel Dan Inscription where we had access to
the artefacts themselves, many conclusions were reached before a thor-
ough (rather than preliminary) investigation of the actual fragments was
made. As a result, the fragments have been displayed at the Israel Museum
in Jerusalem in the wrong configuration for a number of years now. As
tempting as it may be, scholars must put off conclusions before a thorough
analysis is carried out, or else future research taking into account these
premature conclusions will be fundamentally flawed.
Second, scholars should be warned about preconceptions when it comes
to the biblical texts. The historicity of the United Monarchy has been
steadily eroded by the lack of supporting evidence from excavations. Yet
we must question what it actually is we are looking for in excavations to
inform us about the biblical texts concerning David, Solomon and Jerusa-
lem, if indeed it is even a question that should be asked. The extensive
conclusions of Jamieson-Drake1 against a Davidic empire signalled the
death knell for the United Monarchy. However, it by no means banished
the persons of David or Solomon into the ether of mythical characters.
Rather, it appears that for too long we have been understanding the term
'United Monarchy' from a modern European perspective that inherently
requires state entities to possess land and officialdom. However, upon
closer inspection of the biblical texts, it becomes clear that the texts them-
selves do not press for such an understanding. We have, therefore, been
misunderstanding the biblical texts or have trained our sights too high. It
may well be argued that when it comes to the 'United Monarchy' terms
Abou-Assaf, A., P. Bordreuil and A.R. Millard, La statue de Tell Fekherye et son inscription
bilingue assyro-arameenne (Recherche sur les civilisations, 7; Paris: ADPF, 1982).
Aharoni, Y., The Land of the Bible: A Historical Geography (London: Westminster Press, 2nd
rev. edn, 1979).
Aharoni, Y., and M. Avi-Yonah, The Macmillan Bible Atlas (New York: Macmillan, 3rd rev.
edn, 1993).
Ahituv, S., 'Suzerain or Vassal? Notes on the Aramaic Inscription from Tel Dan', IEJ 43
(1993), pp. 246-47.
Ahlstrom, G., The History of Ancient Palestine from the Palaeolithic Period to Alexander's
Conquest (JSOTSup, 146; Sheffield: JSOT Press, 1993).
Albright, W.F., 'New Light on the Early History of Phoenician Colonization', BASOR 83
(1941), pp. 14-22.
Archaeology and the Religion of Israel (Baltimore: The Johns Hopkins University Press,
2nd edn, 1946).
Alfrink, B., 'L'Expression rniDK Dtf 332)', OTS2 (1943), pp. 106-18.
Andersen, F.I., 'I Have Called You By Name...', BH34 (1998), pp. 37-52.
Arav, R., 'Bethsaida 1996-1998', IEJ49 (1999), pp. 128-36.
Aufrecht, W.E., A Corpus of Ammonite Inscriptions (Lewiston, NY: Edwin Mellen Press,
1989).
Avigad, N., 'The Epitaph of a Royal Steward from Siloam Village', IEJ3 (1953), pp. 137-52.
'An Inscribed Bowl from Dan', PEQ 100 (1968), pp. 42-44.
'Excavations in the Jewish Quarter of the Old City of Jerusalem, 1969/1970', IEJ20 (1970),
pp. 1-8, 129-40.
'Excavations in the Jewish Quarter of the Old City of Jerusalem, 1971', IEJ22 (1972), pp.
193-200.
Discovering Jerusalem (Nashville: Thomas Nelson, 1983).
Barstad, H.M., and B. Becking, 'Does the Stele from Tel-Dan Refer to a Deity Dod?', 57V 77
(1995), pp. 5-12.
Becking, B., The Second Danite Inscription: Some Remarks', 5/V81 (1996), pp. 21-29.
'Did Jehu Write the Tel Dan Inscription?', SJOT13 (1999), pp. 187-201.
Bernett, M., and O. Keel, Mond, Stier undKult am Stadttor: Die Stele von Betsaida [et-Tell]
Unter Mitarbeit von Stefan Munger (Gottingen: Vandenhoeck & Ruprecht, 1998).
Biran, A., Biblical Dan (Jerusalem: Israel Exploration Society, 1994).
'Two Bronze Plaques and the Hussot of Dan', IEJ 49 (1999), pp. 43-54.
Biran, A., and J. Naveh, 'An Aramaic Stele Fragment from Tel Dan', IEJ43 (1993), pp. 81-
98.
'The Tel Dan Inscription: A New Fragment', IEJ 45 (1995), pp. 1-18.
Birnbaum, S.A., The Hebrew Scripts. II. The Plates (2 vols.; London: Palaeographia, 1954-57).
The Hebrew Scripts, Part One: The Text (Leiden: E.J. Brill, 1971).
Bibliograph++++++++++++++++++++
Chapman, R.L., 'The Dan Stele and the Chronology of Levantine Iron Age Stratigraphy',
Bulletin of the Anglo-Israel Archaeological Society 13 (1993-94), pp. 23-29.
Coogan, M.D., Stories from Ancient Canaan (Philadelphia: Westminster Press, 1978).
Cook, S.A., 'Inscribed Hebrew Objects from Ophel', PEFQS 56 (1924), pp. 183-86.
Cowley, A., Aramaic Papyri of the Fifth Century EC (Oxford: Clarendon Press, 1923).
Cross, P.M., 'The Origin and Early Evolution of the Alphabet', 78 (1967), pp. 8-24.
'The Stele Dedicated to Melcarth by Ben-Hadad of Damascus', BASOR 205 (1972), pp.
36-42.
'Leaves from an Epigraphist's Notebook', CBQ 36 (1974), pp. 486-94.
'Early Alphabetic Scripts', in idem (ed.), Symposia Celebrating the Seventy-Fifth Anniver-
sary of the Founding of the American Schools of Oriental Research, I (2 vols.; Cam-
bridge, MA: American Schools of Oriental Research, 1979), pp. 97-123.
'Newly Found Inscriptions in Old Canaanite and Early Phoenician Script', BASOR 238
(1980), pp. 1-20.
'King Hezekiah's Seal Bears Phoenician Imagery', BARev 25.2 (1999), pp. 42-45, 60.
Cryer, F.H., 'On the Recently-Discovered "House of David" Inscription', SJOT8 (1994), pp.
3-19.
'A "Betdawd" Miscellany: Dwd, Dwd' or Dwdh?', SJOT+(1995), pp. 52-58+ .
'King Hadad', SJOT9+(1995), pp. 223-35.
'Of Epistemology, Northwest-Semitic Epigraphy and Irony: The "bytdwd/House of David"
Inscription Revisited', JSOT69 (1996), pp. 3-17.
Davies, P.R., 'BYTDWD and SWKT DWYD: A Comparison', JSOT64 (1994), pp. 23-24.
'"House of David" Built on Sand: The Sins of the Biblical Maximizers', BARev 20.4
(1994), pp. 54-55.
In Search of 'Ancient Israel' (JSOTSup, 148; Sheffield: Sheffield Academic Press, 1995).
'Method and Madness: Some Remarks on Doing History with the Bible', JBL 114 (1995),
pp.699-705.
Dearman, A. (ed.), Studies in the Mesha Inscription andMoab (Atlanta: Scholars Press, 1989).
DeCaen, V., 'The Morphosyntactic Argument for the waw Consecutive in Old Aramaic', VT
51 (2001), pp. 381-85.
Degen, R., Altaramdische Grammatik der Inschriften des 10.-8. Jh. v. Chr. (Wiesbaden:
Kommissionsverlag F. Steiner [fur die] Deutsche Morgenlandische Gesellschaft, 1969).
De Moor, J.C., The Rise ofYahwism (Leuven: Leuven University Press, 1990).
Demsky, A., 'On Reading Ancient Inscriptions: The Monumental Aramaic Stele Fragment
from Tel Dan\ JANES 23 (1995), pp. 29-35.
Deutsch, R., and M. Heltzer, New Epigraphic Evidence from the Biblical Period (Tel Aviv:
Archaeological Center Publications, 1995).
Dietrich, M., O. Loretz and J. Sanmartin, Die keilalphabetischen Texte aus Ugarit (Neukirchen-
Vluyn: Neukirchener Verlag, 1976).
Dijkstra, M., 'An Epigraphic and Historical Note on the Stela of Tel Dan', BN 74 (1994), pp.
10-14.
Diringer, D., Le iscrizioni antico-ebraiche Palestinesi (Florence: Le Monnier, 1934).
Donner, H., and W. Rollig, Kanaandische und Aramdische Inschriften(3 vols.; Wiesbaden:
Otto Harrassowitz, 1964).
Ehrlich, C.S., 'The BYTDWD-lnscnplion and Israelite Historiography: Taking Stock after Haifa
Decade of Research', in P.M. Michele Daviau, J.W. Wevers and M. Weigl (eds.), The
World of the Arameans. II. Studies in History and Archaeology in Honour of Paul-Eugene
Dion (3 vols.; JSOTSup, 325; Sheffield: Sheffield Academic Press, 2001), pp. 57-71.
322 The Tel Dan Inscription
Emerton, J.A., 'New Evidence for the Use of Waw Consecutive in Aramaic', VT44 (1994), pp.
255-58.
Eph'al, I., and J. Naveh, 'Hazael's Booty Inscriptions', IEJ39 (1989), pp. 192-200.
Finkelstein, I., 'The Archaeology of the United Monarchy: An Alternative View', Levant 28
(1996), pp. 177-87.
Fitzmyer, J.A., The Aramaic Inscriptions ofSefire (Rome: Pontifical Biblical Institute, rev.
edn, 1995).
Galil, G., 'A Re-Arrangement of the Fragments of the Tel Dan Inscription and the Relations
between Israel and Aram', PEQ 133 (2001), pp. 16-21.
Galling, K., 'The Scepter of Wisdom: A Note on the Gold Sheath ofZenjirli and Ecclesiastes
12:11', BASOR 119 (1950), pp. 15-18.
Garbini, G., 'L'iscrizione aramaica di Tel Dan', Atti della Accademia nazionale del Lincei,
Scienze morali, storiche efilologiche, rendiconti 9.5.3 (1994), pp. 461-71.
Garr, W.R., Dialect Geography of Syria-Palestine, 1000-586 BCE (Philadelphia: University of
Pennsylvania Press, 1985).
Gaus, U., V. Von Grave and M. Kerschner, 'Milet 1990: Vorbericht iiber die Arbeiten des Jahres
1990',/s/Mtf 41 (1991), pp. 125-86.
Gibson, J.C.L., Textbook of Syrian Semitic Inscriptions (3 vols.; Oxford: Clarendon Press,
1971).
Ginsberg, H.L., 'MMST and MSFT, BASOR 109 (1948), pp. 20-21.
Gitin, S., T. Dothan and J. Naveh, 'A Royal Dedicatory Inscription from Ekron', IEJ 47
(1997), pp. 1-16.
Grant, E., and G.E. Wright, Ain Shems Excavations. V. Text & Plates (5 vols.; Bible and
Kindred Studies, 8; Haverford: Haverford College, 1939).
Greenfield, J.C., 'Studies in West Semitic Inscriptions. I. Stylistic Aspects of the Sefire Treaty
Inscriptions', AcOr 29 (1965), pp. 1-18.
Greenfield, J.C., and A. Shaffer, 'Notes on the Akkadian-Aramaic Bilingual Statue from Tell
Fekherye', Iraq 65 (1983), pp. 109-16.
Hackert, J.A., The Balaam Text from Deir 'Alia (Chico, CA: Scholars Press, 1980).
Halpern, B., 'The Stela from Dan: Epigraphic and Historical Considerations', BASOR 296
(1994), pp. 63-80.
Harris, W.V., Ancient Literacy (Cambridge, MA: Harvard University Press, 1989).
Killers, D.R., and E. Cussini, Palmyrene Aramaic Texts (Baltimore: The Johns Hopkins Uni-
versity Press, 1996).
Hoftijzer, J., and K. Jongeling, Dictionary of the North-West Semitic Inscriptions (2 vols.;
Leiden: E.J. Brill, 1995).
Hoftijzer, J., and G. Van der Kooij (eds.), Aramaic Texts From Deir 'Alia (Leiden: E.J. Brill,
1976).
Horn, S.H., 'The Amman Citadel Inscription', BASOR 193 (1969), pp. 2-19.
Ingholt, H., 'Un Nouveau thiase a Palmyre', Syria 1 (1926), pp. 128-41.
Rapportpreliminaire sur la premiere campagne desfouilles a Hama (Copenhagen: Levin &
Munksgaard, 1934).
Jalabert, L., and R. Mouterde, Inscriptions grecques et latines de la Syrie (Paris: Paul Geuthner,
1929).
Jamieson-Drake, D.W., Scribes and Schools in Monarchic Judah: A Socio-Archaeological
Approach (JSOTSup, 155; The Social World of Biblical Antiquity, 9; Sheffield: Almond
Press, 1991).
Bibliography 323
Jaros, K., Hundert Inschriften aus Kanaan und Israel (Fribourg: Verlag Schweizerisches
Katholisches Bibelwerk, 1982).
Jastrow, M., A Dictionary of the Targumim: The Talmud Babli and Yerushalmi, and the
Midrashic Literature (2 vols.; New York: Judaica Press, 1993).
Jepsen, A., 'Israel und Damaskus', AfO 14 (1941^5), pp. 153-72.
Joiion, P., and T. Muraoka, A Grammar of Biblical Hebrew (2 vols.; Rome: Pontifical Biblical
Institute, 1993).
Kallai, Z., 'The King of Israel and the House of David', IEJ43 (1993), p. 248.
Kaufman, S.A., 'Reflections on the Assyrian-Aramaic Bilingual from Tell Fakhariyeh',
Maarav3(1982),pp. 137-75.
Kitchen, K.A., 'A Possible Mention of David in the Late Tenth Century BCE, and Deity *Dod
as Dead as the Dodo?', JSOT 16 (1997), pp. 29-44.
Knauf, E.A., A. De Pury and T.R. Romer, ' *BaytDawld ou *BaytDodT, BN12 (1994), pp
60-69.
Kuan, J.K.J., Neo-Assyrian Historical Inscriptions and Syria-Palestine (Hong Kong: Alliance
Bible Seminary, 1995).
Lapp, P.W., and N.L. Lapp, Discoveries in the Wadi ed-Ddliyeh (Cambridge, MA: American
Schools of Oriental Research, 1974).
Lehmann, R.G., and M. Reichel, 'DOD und ASIMA in Tell Dan', BN77 (1995), pp. 29-31.
Lemaire, A., 'Les Inscriptions de Khirbet el-Q6m', RB 84 (1977), pp. 595-608.
'"House of David" Restored in Moabite Inscription', BARev 20.3 (1994), pp.
30-37.
'The Tel Dan Stela as a Piece of Royal Historiography', JSOT 81 (1998), pp. 3-14.
Lemaire, A., and J.-M. Durand, Les inscriptions arameennes de Sfire et I 'Assyrie de Shamshi-
Ilu (Geneva: Droz, 1984).
Lemche, N.P., 'Is it Still Possible to Write a History of Ancient Israel?', SJOT 8 (1994), pp.
165-90.
'Bemerkungen iiber einen Paradigmenwechsel aus Anlass einer neuentdecken Inschrift', in
M. Weippert and S. Timm (eds.), Meilenstein. Festgabefur Herbert Donner zum 16
Februar 1995 (Wiesbaden: Otto Harrassowitz, 1995), pp. 99-108.
Lemche, N.P., and T.L. Thompson, 'Did Biran Kill David? The Bible in the Light of Archae-
ology', JSOT'64 (1994), pp. 3-22.
Lidzbarski, M., Handbuch der Nordsemitischen Epigraphik nebst A usgewdhlten Inschriften (2
vols.; Hildesheim: Olms, 1962).
Lipiriski, E., 'Le Ben-Hadad II de la Bible et 1'histoire', in P. Peli (ed.), Proceedings of the
Fifth World Congress of Jewish Studies (4 vols.; Jerusalem: World Union of Jewish
Studies, 1969), I, pp. 157-73.
'Aram et Israel du Xe au VIII6 siecle av. N.E.\Acta Antiquita 27 (1979), pp. 49-102.
Luckenbill, D.D., Ancient Records of Assyria and Babylonia (New York: Greenwood, 1927).
Luschan, F. von, Ausgrabungen in Sendschirli. V. Die Kleinfunde von Sendschirli: Herausgabe
undErgdnzung besorgtvon Walter Andrae (5 vols.; Mittheilungen aus den Orientalischen
Sammlungen, Konigliche Museen zu Berlin, 15; Berlin: W. de Gruyter, 1943).
Maisler, B., 'Two Hebrew Ostraca from Tell Qasile', JNES 10 (1951), pp. 265-67.
Margalit, B., 'The Old-Aramaic Inscription of Hazael from Dan', UF26 (1994), pp. 317-20.
Mazar, B., 'Bin Gev: Excavations in 1961', IEJ 14 (1964), pp. 27-29.
Mazar, E., and B. Mazar, Excavations in the South of the Temple Mount: The Ophel of Biblical
Jerusalem (Jerusalem: Institute of Archaeology, Hebrew University of Jerusalem, 1989).
Messerschmidt, L., Keilschrifttexte as Assur: historischen Inhalts (2 vols.; Osnabriick: Otto
Zeller, 1970).
324 The Tel Dan Inscription
Millard, A.R., 'Alphabetic Inscriptions on Ivories from Nimrud', Iraq 24 (1962), pp. 45-49.
Millard, A.R., and H. Tadmor, 'Adad-Nirari III in Syria: Another Stele Fragment and the
Dates of his Campaigns', Iraq 35 (1973), pp. 62-64.
Miller, J.M., 'The Elisha Cycle and the Accounts of the Omride Wars', JBL 85 (1966), pp.
441-54.
'The Fall of the House of Ahab', VT17 (1967), pp. 307-24.
'The Rest of the Acts of Jehoahaz', ZAW 80 (1968), pp. 337-42.
Miller, J.M., and J.H. Hayes, A History of Ancient Israel andJudah (London: SCM Press, 1986).
Muraoka, T., 'Linguistic Notes on the Aramaic Inscription from Tel Dan', IEJ45 (1995), pp.
19-21.
'The Tel Dan Inscription and Aramaic/Hebrew Tenses', Abr-Nahrain 33 (1995), pp. 113-15.
'Again on the Tel Dan Inscription and the Northwest Semitic Verb Tenses', ZAH11 (1998),
pp.74-81.
'The Prefix Conjugation in Circumstantial Clauses in the Tel Dan Inscription?', VT 51
(2001), pp. 389-92.
Muraoka, T., and M. Rogland, 'The waw Consecutive in Old Aramaic? A Rejoinder to Victor
Sasson', FT 48 (1998), pp. 99-104.
Na'aman, N., 'Hezekiah's Fortified Cities and the LMLK Stamps', BASOR 261 (1986), pp. 5-
21.
'Hazael of 'Amqi and Hadadezer of Beth-rehob', UF 27 (1995), pp. 381-94.
'The Contribution of Royal Inscriptions for a Re-Evaluation of the Book of Kings as a
Historical Source', JSOT82 (1998), pp. 3-17.
Naveh, J., 'The Scripts in Palestine and Transjordan in the Iron Age', in J.H. Sanders (ed.),
Near Eastern Archaeology in the Twentieth Century, Essays in Honor of Nelson Glueck
(Garden City, NY: Doubleday, 1970), pp. 277-83.
Early History of the Alphabet: An Introduction to West Semitic Epigraphy and
Paleaography (Jerusalem: Magnes Press, 1982).
Noll, K.L., 'The God Who is Among the Danites', JSOT80 (1998), pp. 3-23.
Parpola, S., and K. Watanabe, Neo-Assyrian Treaties and Loyalty Oaths (State Archives of
Assyria, 2; Helsinki: Helsinki University Press, 1988).
Pitard, W.T., Ancient Damascus: A Historical Study of the Syrian City-State from the Earliest
Times until its Fall to the Assyrians in 732 BCE (Winona Lake, IN: Eisenbrauns, 1987).
'The Identity of the Bir-Hadad of the Melqart Stela', BASOR 272 (1988), pp. 3-22.
Pritchard, J.B. (ed.), The Ancient Near East in Pictures (Princeton, NJ: Princeton University
Press, 1954).
The Times Concise Atlas of the Bible (London: Times Books, 1991).
Puech, E., 'La Stele Arameene de Dan: Bar Hadad II et la coalition des Omrides et de la
maison de David', RB 101-102 (1994), pp. 215-41.
Rainey, A., 'The "House of David" and the House of the Deconstructionists', BARev 20.6
(1994), p. 47.
Reed, W.L., and F.V. Winnett, 'A Fragment of an Early Moabite Inscription from Kerak',
BASOR 172 (1963), pp. 1-9.
Reisner, G.A., C.S. Fischer and D.G. Lyons, Harvard Excavations at Samaria (2 vols.;
Cambridge, MA: Harvard University Press, 1924, 1964).
Rendsburg, G.A., 'On the Writing TnjTD in the Aramaic Inscription from Tel Dan', IEJ45
(1995), pp. 22-25.
Rosenthal, F., A Grammar of Biblical Aramaic (Wiesbaden: Otto Harrassowitz, 1963).
Sanders, J.H. (ed.), Near Eastern Archaeology in the Twentieth Century: Essays in Honor of
Nelson Glueck (Garden City, NY: Doubleday, 1970).
Bibliography 325
Sasson, V., 'The Old Aramaic Inscription from Tell Dan: Philological, Literary, and Historical
Aspects', JSS 40.1 (1995), pp. 11-30.
'Murderers, Usurpers, or What? Hazael, Jehu, and the Tell Dan Old Aramaic Inscription',
C/F28(1996), pp. 547-54.
'Some Observations on the Use and Original Purpose of the Wow Consecutive in Old
Aramaic and Biblical Hebrew', VT 41 (1997), pp. 111-27.
Schniedewind, W.M., 'Tel Dan Stela: New Light on Aramaic and Jehu's Revolt', BASOR 302
(1996), pp. 75-90.
Schniedewind, W.M., and B. Zuckerman, 'A Possible Reconstruction of the Name of Haza'el's
Father in the Tel Dan Inscription', 7E751.1 (2001), pp. 88-91.
Segert, S., 'Zur Schrift und Orthographic der altaramaischen Stelen von Sfire', ArOr 32
(1964), pp. 110-26.
Altaramdische Grammatik mil Bibliographic, Chrestomathie und Glossar (Leipzig: Verlag
Enzyklopadie, 1975).
Seyrig, H., 'Altar Dedicated to Zeus Betylos', in P.V.C. Baur, M.I. Rostovtzeff and A.R.
Bellinger (eds.), Excavations at Dura-Europos Conducted by Yale University and the
French Academy of Inscriptions and Letters: Preliminary Reports of Fourth Season
(New Haven: Yale University Press, 1933), pp. 68-71.
Spycket, A., 'La statue bilingue de Tell Fekherye', RA 19 (1985), pp. 67-68.
Tadmor, H., 'Historical Implications of the Correct Rendering of Akkadian "daku" \JNES 17
(1958), pp. 129-41.
Thiele, E.R., The Mysterious Numbers of the Hebrew Kings (Grand Rapids: Zondervan, rev.
edn, 1983).
Thompson, T.L., 'Dissonance and Disconnections: Notes on the BYTDWD and HMLK.HDD
Fragments from Tel Dan', SJOT9 (1995), pp. 236-40.
' "House of David": An Eponymic Referent to Yahweh as Godfather', SJOT9 (1995), pp.
59-74.
Thureau-Dangin, F. et al, Arslan Task (Paris: Paul Geuthner, 1931).
Torczyner, H., Lachish. I. The Lachish Letters (4 vols.; The Wellcome Archaeological Research
Expedition to the Near East, 1; London: Oxford University Press, 1938).
Tropper, J., 'Eine altaramaische Steleninschrift aus Dan', UF25 (1993), pp. 395-406.
'Palaographische und linguistische Anmerkungen zur Steleninschrift aus Dan', UF 26
(1994), pp. 487-92.
'Aramaisches wyqtl und hebraisches wayyiqtoV, UF 28 (1997), pp. 633-45.
Tufnell, O., M.A. Murray and D. Diringer (eds.), Lachish. III. The Iron Age (4 vols.; The
Wellcome-Marston Archaeological Research Expedition to the Near East, 3; London:
Oxford University Press, 1953).
Tushingham, A.D., 'The Western Hill under the Monarchy', ZDPV95 (1979), pp. 39-55.
Ugolini, M., 'II dio (di) pietra', Sandalion 4 (1981), pp. 7-29.
Ussishkin, D., 'Royal Judean Storage Jars and Private Seal Impressions', BASOR 223 (1976),
pp. 1-13.
'The Destruction of Lachish by Sennacherib and the Dating of the Royal Judean Storage
Jars', Tel Aviv 4 (1977), pp. 28-60.
Van der Toorn, K., 'Herem-Bethel and Elephantine Oath Procedure', ZAW98 (1986), pp. 282-
85.
'Anat-Yahu, Some Other Deities, and the Jews of Elephantine', Numen 39 (1992), pp. 80-
101.
Van der Toorn, K., B. Becking and P.W. Van der Horst (eds.), Dictionary of Deities and
Demons in the Bible (Leiden: E.J. Brill, 1995).
326 The Tel Dan Inscription
BIBLE