Chapter 1<\/p>\n
INTRODUCTION<\/p>\n
Introductory Remarks<\/p>\n
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.
\nHowever, 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 fragments. To date, no one has attempted a thorough and definitive analysis of the fragments. I hope that this study will fulfil that need.
\nFor 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 fragment discovered is labelled Fragment A, the second is labelled Fragment B1, and the third is labelled Fragment B2.<\/p>\n
Methodology<\/p>\n
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 special reference to the Tel Dan fragments. Demsky\u2019s article was mostly written before the publication of Fragments B1 and B2 with only an addendum 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:<\/p>\n
1. To establish the archaeological context in which the inscription was found.
\n2. 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.
\n3. A palaeographical analysis.
\n4. A linguistic analysis.
\n5. An historical synthesis.<\/p>\n
While I agree with the general outline of Demsky\u2019s approach, I believe there is room for improvement. Since the Tel Dan fragments have idiosyncratic 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:<\/p>\n
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 location in which each fragment was found and determining a date for the immediate context through the witness of pottery assemblages 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.
\n2. 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 between the three fragments with regards to their epigraphical characteristics.
\n3. To conduct a palaeographical analysis of the script on each fragment. This will involve noting the shapes of each grapheme on each fragment and comparing them to those of other relevant Northwest Semitic inscriptions, as well as to each other. The purpose of this will be to determine what influences were exerted over the script of each fragment, to try to assign a chronological bracket for the script of each fragment, and to determine whether there is any relationship between the three fragments.
\n4. To produce a synthesis of the epigraphical and palaeographical information gleaned from each fragment. The purpose of this will ultimately be to determine whether there is any relationship between any of the three fragments. In turn, this will also provide a model by which to critique the arrangement of the fragments currently proposed by the publishers, Biran and Naveh, and in which the fragments are currently displayed at the Israel Museum in Jerusalem. If any relationship is determined to exist between any of the fragments, it will be possible to test to see if an arrangement of those fragments can be proposed on the available information. This will provide a framework with which to understand the text(s) properly.
\n5. To undertake a textual analysis of the text(s). This will consider grammatical, syntactical, philological and contextual considerations of the text(s) in the arrangement(s) determined by the previous steps.
\n6. To provide a historical synthesis of the information gleaned from all previous steps. This will involve interpreting the text(s) and attempting to reconstruct the history behind the content, production and destruction of the text(s). In the course of this reconstruction, the date(s) and author(s) of the text(s) will be determined as precisely as the data allows. Inherent in this process will be the incorporation of external data, namely, artifactual and textual sources. It is important that this synthesis with external data is reserved for the final step so as not to prejudice preliminary conclusions.
\n7. To provide a comment on key issues highlighted by the overall synthesis of the available information.<\/p>\n
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 conclusions from one of these steps will often be interspersed with those of the other.
\nIt 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 maintained. 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.
\nFor 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 confident that this study will provide a coherent reappraisal and reinterpretation of the Tel Dan fragments.<\/p>\n
Chapter 2<\/p>\n
ARCHAEOLOGICAL CONTEXT OF THE FRAGMENTS<\/p>\n
Introductory Remarks<\/p>\n
For any inscriptional fragment it is important to first establish its archaeological 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.
\nFor 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 individual strata. The composition of structural walls will also aid us in discerning phases. This will provide us with a chronological bracket in which to place the fragments.
\nAll 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 determines dates for Tel Dan\u2019s stratigraphy, rather than working the other way around in a far safer and more methodical way. Chapman\u2019s reversed approach even leads him to reformulate the chronology of Iron II pottery. The dangers of pursuing such radical conclusions, or even more mundane theories, before the archaeological context of an inscription is thoroughly investigated is to \u2018jump the gun\u2019, so to speak. This can easily lead to disqualification of one\u2019s arguments. In this regard, the comments of Lemche and Thompson regarding the \u2018premature\u2019 appraisal of Fragment A by Biran and Naveh and its subsequent distorting effect find some relevance. In fairness to Biran and Naveh, they did certainly try to present a comprehensive analysis of the archaeological context for all three fragments. However, in the euphoria which surrounded the find of the fragments, it was not realized that the archaeological context of a fragment can only really be fully understood once the entire area of the find site has been excavated and the stratigraphy, pottery and structural remains thoroughly examined. Thus, much of the scholarship surrounding the Tel Dan fragments went ahead on shaky foundations.
\nHalpern was the first to notice that the context in which Fragment A was found might not be as simple as had been first reported. Many of his observations are pertinent to the present study. His keen observations also set the tone for examining the context of Fragments B1 and B2 also, though his study predated the publication of those fragments and so could not deal with them. Although all three fragments were found in the same general area of Tel Dan, analysis must proceed by examining the context in which each individual fragment was found.<\/p>\n
Fragment A<\/p>\n
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 \u2018A\u2019 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.
\nAlong 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 threshold at the base of the perimeter wall. 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 ma\u1e63\u1e63ebah (sacred stones; sg. ma\u1e63\u1e63ebah). Legal proceedings 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 ofthe tel.<\/p>\n
Figure 2.1. Fragment A (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z.Radovan).<\/p>\n
Figure 2.2. Plan of Structure A at the Tel Dan gate complex, indicating the locations at which the Tel Dan fragments were uncovered (Tel Dan Excavations, Hebrew Union College, Jerusalem).<\/p>\n
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 \u2018builders reused a piece of it in the paving of the piazza.\u2019 This confusion over whether the fragment was found at the base of a wall or as part of the paving of the piazza attracted much criticism from leading scholars. Any doubt was dispelled with the publication of Fragments B1 and B2 in which Biran and Naveh unequivocally stated that Fragment A was found \u2018in a wall built on the flagstone pavement of the square\u2019.
\nThe location of Fragment A, however, was not the only point that attracted controversy. Biran and Naveh\u2019s dating of the fragment also unleashed criticism. According to Biran and Naveh, the layer of debris covering the wall in which Fragment A was found was datable to TiglathPileser III\u2019s conquest of northern Israel in 733 BCE. This provided a sound terminus ad quem 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. 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 Cryer and confirmed by later reports of the complex east of the piazza.
\nCryer\u2019s 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.
\nThese observations by Cryer do indeed call into question the dating of the fragment\u2019s 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\u2019s east is slightly more complex than the equation drawn by Cryer.
\nA 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. 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. 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-documented invasion of the Galilee by Tiglath-Pileser III in 733 BCE, which seems to have informed Lemche and Thompson\u2019s conclusion. Three sources attest to this invasion\u2014namely, the destruction layers in Upper Galilean sites (such as Hazor and Tel Dan) and Transjordanian sites (such as Ramoth-Gilead), the Assyrian texts of Tiglath-Pileser III and the biblical text of 2 Kgs 15:29. This invasion, evidenced at Tel Dan by a layer of debris over the gate complex, provides an anchor point for the pottery at Tel Dan.
\nIn addition, Halpern\u2019s initial observations (see below) and Biran\u2019s later excavations (see below also) highlight a number of phases in the strata beneath this layer of debris. The nature of these phases allows us to gain a cumulative picture of the structural development at Dan\u2019s gate. The pottery associated with these phases, therefore, can be accurately placed in a chronological order according to the order in which their associated structures were built. This order, in turn, can be confirmed by the geopolitical climate of their respective eras, known from textual sources. This chronological progression of construction phases is augmented by Finkelstein\u2019s recent redating of the Iron I-II pottery assemblages to a \u2018low chronology\u2019. The construction phases in question at Tel Dan date to the very end of the period which Finkelstein wishes to readjust and are, therefore, relatively unaffected by his chronological shift. Only some slight readjustment of the dates downward is necessary with regards to these particular construction phases at Dan\u2019s gates. Thus, there is very little, if any, ambiguity about the pottery assemblage from Area A of Tel Dan. Noll also realized this important factor of Finkelstein\u2019s chronological adjustment. However, his treatment of the archaeological context for the Tel Dan fragments preceded Biran\u2019s more detailed report of the construction phases evident in the structure to the piazza\u2019s immediate east. As such, Noll\u2019s dating for the archaeological context of the fragments remained statically dependent on Biran and Naveh\u2019s report of the archaeological context as described in the initial publications of the fragments. The thrust of his argument, however, was certainly in the right direction.
\nHalpern drew attention to the differing natures of the various walls abutting the paved piazza, noting that there was \u2018no direct stratigraphic connection between the wall into which the fragment [Fragment A] was integrated and the inner gate structure itself.\u2019 As Halpern observed, the wall in which Fragment A was used (W.5073) appears to have been constructed differently from those nearby. This wall measured c. 2.5 m in thickness and was mostly composed of large stones. In contrast, the more northerly portions of this same wall are considerably thinner, measuring c. 1 m in thickness, and are composed of much smaller stones. The same observations may be made for an adjoining wall that lies perpendicular to the northern extremity of this first wall. This perpendicular wall lies c. 4 m south of the main city wall. Thus, Halpern was correct in his assertion that these thinner portions of the building on the piazza\u2019s eastern edge were certainly not meant to be fortifications. They are far too weak to have served such a purpose. As such, they cannot have been on the edge of the gate complex. Indeed, unless the gap between the main city wall and the building complex on the piazza\u2019s perimeter was blocked off somehow, the entire gate complex would have been rendered useless for both defense and control of traffic. Also, the wall with Fragment A in it differed from most nearby walls in its essential composition. By virtue of these facts, Halpern correctly concluded that there were \u2018at least two, and possibly several, stages in the construction of the structure as a whole\u2019.
\nFollowing publication of all the Tel Dan fragments, further excavation of the building complex to the piazza\u2019s east was undertaken and a brief report was published in Israel Exploration Journal by Biran. The excavations revealed that this building complex had undergone three distinct building phases. The earliest phase consisted of a large rectangular room (5 \u00d7 10.5 m) with a smaller room on the east (3 \u00d7 5 m). The pottery found on the floor of this structure dated to the first quarter of the ninth century BCE, while the pottery beneath the floor was dated to the end of the tenth century BCE. If we take Finkelstein\u2019s 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 \u2018Structure C,\u2019 is therefore datable to approximately the second quarter of the ninth century BCE, contemporary with the Omride Dynasty. 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, a likely scenario since this structure predates most of the gate complex in Area A.
\nThe 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, 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, augmenting the date for this second phase, termed \u2018Structure B\u2019. Finkelstein\u2019s 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\u2019s reign.
\nThe 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 \u2018Structure A\u2019. Cryer\u2019s observation that the stone\u2019s solid composition and flat configuration led to its usage as a building material is quite pertinent, 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, a date unaffected by Finkelstein\u2019s low chronology. This means that the original 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.
\nWe 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 centuryBCE, but before the third construction phase, represented by Structure A, in the early eighth centuryBCE. 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.<\/p>\n
Fragment B1<\/p>\n
Fragment B1 (Fig. 2.3) was found on 20 June 1994, during excavations along the base of Dan\u2019s 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 \u00d7 2.5 m in dimension. This platform had served as a shrine to three ma\u1e63\u1e63ebot found on the northern edge of the platform. The three ma\u1e63\u1e63ebot were arranged along this edge from left to right, largest to smallest. Directly in front of the largest ma\u1e63\u1e63ebot 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 confirmed the platform\u2019s use as a cultic shrine at which offerings were made.
\nThis small shrine was built over debris dated to Tiglath-Pileser Ill\u2019s 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. Fragment B1 was discovered approximately 2 m south of the shrine\u2019s platform (location \u2018B1\u2019 in Fig. 2.2). The fact that Fragment B1 was found in this debris suggests that it came from a later context than that of Fragment A, which had been covered by this debris some 13 m to the southwest. However, this view had to be altered with the discovery of Fragment B2 just ten days later.<\/p>\n
Figure 2.3. Fragment B1 (Tel Dan Excavations, Hebrew Union College, Jerusalem: photograph: Z.Radovan).<\/p>\n
Fragment B2<\/p>\n
As excavations along the base of the Iron Age city wall continued eastward, five ma\u1e63\u1e63ebot were uncovered at the base of the wall. A probe beneath the two easternmost ma\u1e63\u1e63ebot revealed that they had been laid prior to the paving of the area with flagstones. The base of the ma\u1e63\u1e63ebot 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. Therefore, 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 contemporary 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).
\nOn 30 June 1994, Fragment B2 (Fig. 2.4) was discovered as a constituent part of this pavement, contemporary with Structure A. The builders of this pavement had used Fragment B2 as a flagstone where the pavement met the city wall\u2014approximately 8 m north of where Fragment B1 had been found, and just east of the shrine platform dated to after 733 BCE (location \u2018B2\u2019 in Fig. 2.2).
\nThe fact that Fragment B2 was found in this paved area makes it contemporary with Fragment A, as both these fragments were recycled as building material during the same phase of construction in the early eighth century BCE. Thus, both fragments had been on monumental display just prior to this construction phase.<\/p>\n
Figuer 2.4. Fragment B2 (Tel Dan Excavations, Hebrew Union College,Jerusalem; photograph: Z. Radovan).<\/p>\n
Synthesis of the Archaeological Data<\/p>\n
Although Fragment B1 was found in a later context than the other two fragments, we have no choice but to see it as a contemporary of these two other fragments. The reason for this is the unequivocal join between Fragment B1 and Fragment B2. These two fragments fit together so closely and the text on their surfaces match so well that there is absolutely no doubt that they form an integral unit (Fig. 2.5). Since Fragment B2 has the earlier provenance, we must amend our dating for Fragment B1 to match that of Fragment B2. Therefore, the archaeological context in which Fragment B1 was found must be considered tertiary, whereas Fragment A and Fragment B2 were found in secondary contexts. We do not, therefore, know how Fragment B1 was used in its secondary situation. It most probably was recycled as part of a structure that was destroyed in Tiglath-Pileser Ill\u2019s invasion.<\/p>\n
Figure 2.5. Fragment B, comprised of Fragment B1 (top) and Fragment B2 (bottom) joined together (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan).<\/p>\n
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\u2013782 BCE) and Jeroboam II (c. 782\u2013753 BCE). The pottery assemblage found in the levels of this construction phase spanning the first half of the eighth centuryBCE means we should assign this construction phase to the reign of Jehoash (c. 798\u2013782 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 (collectively \u2018Fragment B\u2019) 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.<\/p>\n
Chapter 3<\/p>\n
EPIGRAPHICAL ANALYSIS<\/p>\n
Introductory Remarks<\/p>\n
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.
\nIn this part of my analysis, I will make close and careful observations regarding the physical fragments themselves, and the nature of the individual 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 \u2018Fragment B\u2019.<\/p>\n
Fragment A: Physical Characteristics and the State of Preservation<\/p>\n
Fragment A is made of basalt stone local to the area of Tel Dan. This was confirmed by a mineral analysis conducted by Ariel Heimann very soon after its discovery. Tel Dan itself lies barely a kilometre from the edge of a belt of Neocene-Quaternary basalt, a volcanic rock that makes up most of the regions of the Golan and the eastern Galilee. The actual mound of Tel Dan is located beside the springs of \u2018En Leshem and \u2018En Dan on the eastern bank of Na\u1e25al Dan, the main source of the Jordan River. The basalt, being at such close proximity and possessing good durability, was a natural choice for use in an inscription.
\nThe right-hand edge of Fragment A is the only original edge of the inscription still preserved. The other edges represent breakages of the 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 lacerations suggests that they were sustained during either breakage of the original inscription, transportation, or simply as a result of wear and tear while on display or in secondary usage.
\nA 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 fragment 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 fashioned and smoothed. Of itself, this implies the inscription was of quite some significance.
\nThe fashioned smoothness of the fragment\u2019s 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:<\/p>\n
1. Structural\u2014the 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 structure of a wall; or
\n2. Aesthetic\u2014the inscription was intended to be a free-standing stele, much like that of the Mesha Stele from Dh\u012bb\u0101n in Moab. As such, the stone had to be groomed in order to be aesthetically pleasing from all angles. To this end, not just the written face and right-hand side would have been smoothed, but all sides, including the bottom.<\/p>\n
In order to assess which of these two possibilities is the more probable, we must consider the stone\u2019s state of wear, as well as the wider archaeological context in which it was found. Differing patterns of erosion on the extant faces may betray different levels of exposure to weather and thus hint at how the stone was originally displayed. However, since the fragment was not found in its original display position, we must factor in the effects of breakage, movement and repositioning to the base of the eastern piazza wall where it was uncovered.
\nFirst, the entire vertical length along the right-hand side of Fragment A displays clear signs of gradual and smooth erosion. This belt of erosion extends from the right-hand edge through to the approximate position of the fourth letter of each line, but grows wider towards the bottom of the inscription (see Fig. 3.1). Since there is a consistency to the deteriorated area, it must be concluded that it is the result of natural weathering due to consistent exposure to the elements<\/p>\n
Figure 3.1. Facsimile of Fragment A highlighting area of erosion on the right side of the wtitten surface (drawn by Ada Yardeni).<\/p>\n
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 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.
\nThis 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, suggesting that the original whole inscription was rounded at the top in a similar way to the Mesha Stele. 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 Fragment A shows no signs of this.
\nThe 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 \u2018chisel\u2019, had a round edge, presumably like the end of a matchstick. 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 \u2018stylus\u2019, as Biran terms it, is not an effective 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 experiments conducted by the author, such a tool inevitably produced incisions on stone that were comparable to those on Fragment A.
\nBiran also deduces that the engraving tool was probably made of iron. Certainly the material could not have been flint or stone since such materials are not given to being shaped into long, thin tools which are strong enough to withstand hammering and carving into stone. Tel Dan definitely had a thriving metal industry, as is evidenced by the discovery of numerous crucibles and tuy\u00e9res in Stratum IV of Area B, datable to the middle or late tenth century. However, no iron slag was found in these metalworking installations. Yet, many iron implements, such as sickles, knives and nails were found throughout Strata III, II and I, dating from the ninth to the sixth centuries BCE, making Biran\u2019s suggestion quite plausible indeed.<\/p>\n
Figure 3.2. Paradigmatic cross section of the written surface showing the general roundness of incisions.<\/p>\n
The depth of the incisions never really exceeds 2\u20133 mm, but it varies according to the relative positioning of the incision on the stone. Those incisions made along the belt of erosion on the right-hand side of the written surface are shallower than elsewhere. Thus, they have the appearance of being flatter and having softer edges.
\nIn addition to determining the exact nature of the original inscription and how it was displayed, we must address the claims of forgery labelled against the fragments by Cryer. Since the argument for this claim rests on evidence from the surfaces exposed in breakage, we will need to examine these surfaces as closely as the written surface.<\/p>\n
Fragment A: Epigraphical Analysis<\/p>\n
Mode and Method of Engraving
\nEvery inscribed stroke provides vital information on how the inscription was engraved onto the stone. A shallow point at the extremity of an incision is characteristic of the point at which the engraver started to chisel that stroke. Similarly, a deep point at the other extremity of an incision is characteristic of the point at which the engraver finished incising that stroke. This theory has been borne out by numerous experiments conducted by the author. In Line 1 all the inscribed strokes display clear shallow points at the lowest extremities of the stems, and significantly deeper points at the topmost extremities of the letters (see Fig. 3.3). Evidently, then, the engraver of the inscription incised the letter strokes in an upward direction (see Fig. 3.4). Due to the regularity of this directional information, we can surmise that the engraver was restricted to this upward direction for reasons of comfort and control during engraving.<\/p>\n
Figure 3.3 Deep and shallow points on the incisions indicating the starting point and direction of engraving for the individual letter strokes (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z Radovan [detail]).<\/p>\n
Figure 3.4. Illustration showing the direction in which individual letter strokes were inscribed (datail from facsimile drawn by Ada Yardeni).<\/p>\n
The direction of carving is very important if we are to calculate where the engraver was positioned in relation to the stone when he carved the inscription. Knowing this piece of information will allow us to compare the mode and method of engraving on the fragments and so aid us in determining whether a relationship exists between Fragment A and Fragment B.
\nThe natural posture for a craftsman is to grip the hammer with his dominant hand. Thus, a right-handed craftsman would hold the hammer with his right hand and the chisel with his left; a left-handed craftsman would have the opposite grip. The grip does not pose any restrictions for carving vertical strokes because the engraver can just as easily carve upwards as downwards. All that is needed is a change in the angle at which the chisel is held. For downward strokes, the chisel is angled towards the body, while for upward strokes, the chisel is angled away from the body. Both postures are equally comfortable and afford the same amount of control during engraving because neither posture requires one arm to be stretched across the other. However, for horizontal strokes, the engraver is very much restricted in the direction in which he carves. An engraver will not naturally carve a stroke with a grip that requires his arms to be crossed while he swings the hammer and positions the chisel. Thus, a right-handed craftsman will naturally carve horizontal strokes from right to left, while a left-handed craftsman will naturally carve from left to right.
\nWe expect, then, to find strokes viewed by the engraver as horizontal to be carved in essentially one direction, while vertical strokes to be carved in either of two directions (up or down). Assuming the engraver was right-handed, the directional regularity of the letters in Line 1 allows us to postulate that the engraver was positioned to the left of the inscription. This means that the strokes which we view as vertical when reading the text would have been seen as horizontal by the engraver positioned at the left edge of the inscription.
\nWhen we examine the horizontal stroke on the head of the resh in Line 1, we find a deepening at the extreme right. The tell-tale sign of an actual deep finishing point, however, has been carved over by the stem of the resh, but the evidence nevertheless points to the stroke having been carved from left to right. Again, assuming the engraver was right-handed, it would be extremely difficult to have carved these letters while positioned at the bottom edge of the inscription, which is the most natural position for reading the inscription. Indeed, this would require the engraver\u2019s arms to have been crossed, greatly diminishing both precision of chiselling and control of the hammer. As such, we would not expect the letters to have been carved as neatly as they have been.
\nFurthermore, since the stem of the resh was evidently the final stroke of the letter chiselled, we have further support for advocating the engraver\u2019s position at the left edge of the inscription. The \u2018nose\u2019 of the figure resh (the leftmost apex) appears to have been the starting point for carving the letter. Thus, both strokes forming the \u2018head\u2019 of the letter emanate from this point out towards their respective deep finishing points along the stem, which was evidently the last stroke of the letter engraved. Thus, we observe the tendency to start engraving letters from the leftmost strokes and 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.
\nThis 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 overwhelmingly 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. 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 carving and the order of strokes carved.
\nA fiirther 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 inscription while carving it from the stone\u2019s left edge\u2014a 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, 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.<\/p>\n
Significance of the Slope of Each Textual Line
\nAn analysis of the slope of each textual line in Fragment A bears this theory out. One of the peculiarities of Fragment A is that the first few lines of text appear to slope considerably down towards the left-hand side. However, one notices the lines becoming gradually flatter and more level with the horizontal towards the end of the inscription. Table 3.1 lists the average slope of each line in comparison to the horizontal. The axis is the imaginary line from which each letter is hung, starting at the visible beginning of each line as we have it and taking the right edge of the fragment as being vertical.
\nSince the beginning of Line 1 and Line 2 are no longer extant, the slope of these two lines may appear to be statistically misleading. However, comparison with the same portion of more complete lines beneath these two lines indicates that this portion of each line actually represents the average slope of the line, thus validating their figures. This does, however, mean that the slopes of the last two lines may be statistically misleading since the portion of the slope representing the average is no longer extant on them. Hence, they are shaded in Table 3.1.<\/p>\n
Table 3.1. Slope of each line in Fragment A<\/p>\n
Fragment A
\nDegrees:
\n+ above the horizontal
\n\u2212 below the horizontal
\nLine 1
\n\u221211\u00b0
\nLine 2
\n\u221210\u00b0
\nLine 3
\n\u22128\u00b0
\nLine 4
\n\u22125\u00b0
\nLine 5
\n\u22124\u00b0
\nLine 6
\n\u22123\u00b0
\nLine 7
\n\u22122\u00b0
\nLine 8
\n0\u00b0
\nLine 9
\n0\u00b0
\nLine 10
\n0\u00b0
\nLine 11
\n0\u00b0
\nLine 12
\n+2\u00b0
\nLine 13
\n+4\u00b0<\/p>\n
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.
\nWhen 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.
\nSince 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 letters in the top part of Fragment A are attributable to the process of engraving. 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 readability and comprehension of the text. Second, even if the engraver could have produced the original inscription firom 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 here\u2014there is a certifiable trend for the lines to become more horizontal, and as such, there must have been a consistent factor behind it.
\nAs 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 the final line of the original inscription. Thus, the original inscription would have stood taller, and with more lines of text, than the current sample from Fragment A.
\nWith 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 original 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.
\nThis is also evidenced by the somewhat cramped nature of the letters in these upper portions of Fragment A, and the gradual improvement of general neatness towards the bottom of the firagment. 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.
\nWe should ask whether any other factors could have caused the improvement in the inscription\u2019s 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.
\nIt 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 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.
\nHaving 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 engraver 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\u2019s 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.<\/p>\n
Dimensions of the Original Inscription
\nWe can now tie together all the discernible data in an attempt to calculate the dimensions of the original inscription. Bearing in mind that the stone lay flat during the writing and engraving processes, it must be noted that the decided slant in the first few extant lines tapers off at approximately Line 8. In accord with the theory of the scribe\u2019s posture offered above, it can be deduced that at Line 8 of Fragment A, the scribe\u2019s arm was not overstretched, hence the noted improvement in the fragment\u2019s general neatness. However, for the lines above Line 8, the scribe\u2019s arm appears to have been extended, quite probably with the scribe himself leaning forward over the stone. Assuming that the scribe\u2019s arm was between 60\u201370 cm long, we can surmise that Line 8 of Fragment A was at least 60\u201370 cm above the original bottom edge of the whole inscription. We can average this out to 65 cm.
\nFurthermore, 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\u2019s height above Line 1 to our estimations. These extra 20 cm yield us a minimum approximate height of 85 cm.
\nYet, 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. 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 inscription stone when it stood in tact. Since neither Fragment A nor Fragment B1 show any trace of the inscription\u2019s original top edge, there is no way of suggesting a maximum height for the inscription.<\/p>\n
Figure 3.5. Illustrations demonstrating the scribe\u2019s 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).<\/p>\n
It must be noted, however, that this estimation is for the height of the stone itself\u2014not 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. 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.
\nAs we turn to the task of deriving an approximate width for the inscription 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.
\nFirst, as was noted previously, the individual letter strokes point towards the engraver having been situated along the left edge of the inscription. We 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\u2019s 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\u2019s 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\u2019s posture.
\nHowever, 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 inscription\u2014just a position that affords visibility of individual strokes composing the letters, and comfort in engraving these strokes with a chisel and mallet. Such positions 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).<\/p>\n
Figure 3.6. Illustration showing the position of the engraver in relation to fragment A and the inscription as a whole (not to scale).<\/p>\n
Grip and control of a chisel with the simultaneous swinging of a mallet, whether swung from a high or low position, is best carried out with the elbows slightly bent. If the arms are kept straight, precision of the chisel is lost and the accuracy of aiming at and hitting the chisel with a mallet decreases significantly. It is also natural to lean the forearm or elbow of the arm controlling the chisel on the writing surface, reducing fatigue and improving wrist leverage for guiding the chisel. With such a posture and the considerations that it allowed easy control of carving on the far right edge, the width of the inscription is likely to have been no more than the approximate length between the elbow and fingers of the engraver. This is almost identical to a classic unit of measurement, the cubit, which measured approximately 45 cm. If the width had been greater than this, we would expect the engraver to have carved at least some of the letters, especially those along the right edge, from another edge around the stone\u2019s original perimeter for greater ease. Of this, we have no evidence.
\nThere is one more clue which permits us to refine this approximation even further, and that is the large semi-circular recess along the left edge of Fragment A, between the latitudes of Line 6 and Line 10. Part of this recess is filled with a small wedge-shaped fragment of stone that was found with the entire fragment. I will call this fragment the \u2018micro-fragment\u2019. Oddly enough, the micro-fragment is missing from its place in a photograph which Biran and Naveh published in their first article announcing the find of Fragment A (up to that time, the only fragment discovered). This oddity drew sharp criticism from Cryer at Copenhagen University, who then questioned Biran and Naveh\u2019s reporting of Fragment A\u2019s place within the archaeological context of Tel Dan. What is even more peculiar is that neither Biran nor Naveh, nor indeed any other scholar connected with the debate over the Tel Dan fragments, tendered an explanation for this recess and its significance.
\nDespite this gap in critical evaluation, the nature of the recess and the micro-fragment within it tells us what caused it. First, the recess is semicircular in shape and we may reasonably deduce that the entire recess was roughly circular. Thus, the recess has been artificially made. It is not the result of a natural fissure in the rock or of weathering. Nor indeed could it have resulted from the transference of the fragment to its secondary place 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.
\nSecond, 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 instrument 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.
\nPresumably, 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\u2019s elbow or waist: about 85\u201390 cm. This matches perfectly with the calculated height of Line 8, which intersects the recess.
\nThus, 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 approximate 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 110 \u00d7 35 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\u2019s 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.<\/p>\n
Line 1
\nThe first line extant on Fragment A preserves only four letters. By comparison with better preserved lines (e.g. Line 5), it is calculated that there 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 exposed 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).<\/p>\n
Figure 3.7. Illustration showing enough room for at least two lines of text before Line 1 of Fragment A (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan [detail]).<\/p>\n
Biran and Naveh understood the first visible strokes at the extant beginning of Line 1 as forming the figure of a mem. 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 plausible, 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 \u0161in. 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 \u0161in would be too close together. 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 letters in Line 8 than there is here in Line 1. Thus, there is ample evidence to allow the reconstruction of a taw and \u0161in here.
\nYet, despite the room for this reconstruction, we must test to see if it actually is the case. Schniedewind\u2019s 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 \u0161in 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 \u0161in 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 \u0161in, 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 \u0161in.
\nThis leaves us with the vertical stroke preceding this \u0161in. Tropper\u2019s 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 and peh. A mem is automatically ruled out for lack of space. Similarly, all instances of kaph, nun and peh in Fragment A (and Fragment B) are too large to fit in this narrow space. We have no such problems with accommodating a taw. Therefore, Tropper\u2019s suggestion of a taw followed by \u0161in at the beginning of Line 1 is to be retained.
\nThe incisions of both the taw and \u0161in 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, 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\u2019s prepared surface before the inscription was engraved. Neither can it be the result of an accidental slip of the engraver\u2019s 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.
\nThe 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. 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.<\/p>\n
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]).<\/p>\n
The leftmost chip (\u2018Chip 3\u2019 in Fig. 3.8) provides the vital clue. At the leftmost extremity of this chip, we observe an abrasion that is clearly associated 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 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 conclude that the patch of erosion which covers the first two letters, taw and \u0161in, and a small area around it was not sustained while the inscription was in its original display position, but after the inscription was broken. Subsequently, the three chips were sustained during the transfer and reuse of the fragment as building material.
\nAfter the taw and \u0161in 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.
\nFollowing 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.
\nThe final letter extant on Line 1 is an \u2018ayin. The letter is almost fully preserved with only the very left side of the figure damaged in the inscription\u2019s 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.
\nWe may decipher the inscribed characters on Line 1, therefore, as: [\u2026]\u05d7\u05af\u05e9\u05e8\u00b7\u05e2[\u2026].<\/p>\n
Line 2
\nThe beginning of Line 2 is very difficult to decipher. The broken edge where the written surface meets one of the newly exposed faces (the uppermost) 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 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.<\/p>\n
Figure 3.9. The abraded edge between the written face and the uppermost exposed face (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan [detail]).<\/p>\n
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\u2019s facsimile of Fragment A offers no reconstruction either, though she shows that some unclear markings are present on the face of the stone. 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, \u1e63adhe, 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 \u1e6deth, yodh, lamed and \u2018ayin, each of which must be dealt with on their own merits.
\nThe difficulty in proposing a \u1e6deth here is that neither Fragment A nor Fragment B contain any examples of \u1e6deths. Thus, we have to turn to the evidence from other similar monumental lapidary inscriptions that do attest the letter \u1e6deth. The Amman Citadel Inscription has one instance of \u1e6deth as a right-leaning elipse with a score across its width at the centre. This would indeed be a suitable match for the relevant etched marks on Fragment A. The Sefir\u00e9 Treaties attest a similar left-leaning \u1e6deth. The Mesha Stele attests the letter \u1e6deth as a circle with two interior radii forming a sector inside the circle. The inscription from Tell Fakhariyah attests a similar \u1e6deth as a circle with two intersecting diameters, as do Azitawadda\u2019s inscription from Karatepe and Panammu I. Indeed, from the A\u1e25iram Inscription to the Eshmun\u2018azar Sarcophagus, the letter \u1e6deth is attested as round and therefore a candidate for consideration as the damaged letter here in Line 2.
\nThere are two problems, however, with identifying this letter as a \u1e6deth. First, from the remnant of the damaged letter, there appears to be no evidence of any strokes inside the rounded portions (i.e. the letter\u2019s interior). This means that if we are to understand this letter as a \u1e6deth, the interior stroke characteristic of a \u1e6deth must be similar to the \u1e6deth in Line 10 of the Mesha Stele (in the word \u05e2\u05d8\u05e8\u05ea). This sees the interior stroke as resembling two radii extending from the centre point out to the letter\u2019s 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 \u1e6deth and one that is rare and not consistent even within the Mesha Stele. The other possibility is to consider a shape similar to the \u1e6deth in the Amman Citadel inscription.
\nHowever, 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 \u1e6deth, 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 \u1e6deth would be very oddly placed. For this reason, and the one stated above, we must have serious reservations about suggesting \u1e6deth as the damaged letter here.
\nWe find similar problems with the suggestion of an \u2018ayin being the damaged letter. Since both \u1e6deth and \u2018ayin are of comparable height and shape in other monumental lapidary inscriptions, the placement of the letter is troublesome. Since we do have other attestations of \u2018ayin in Fragment A, we do not need to appeal to other inscriptions for clarification of the letter shape. It is quite clear that the \u2018ayin 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 \u1e6deth.
\nAnother 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\u20138 provide us with good templates. The advantage of suggesting a lamed is that we have no problem with the letter\u2019s 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.
\nSchniedewind also reconstructs this letter as a lamed, but he does so on the basis of \u2018computer enhancement and imaging\u2019 and the assumption that the letters reconstruct a name (as opposed to common noun, verb, and so on). In fact, in his initial transcription of the fragments as he arranges them, Schniedewind deals with this letter as though the letter was not in doubt but only slightly damaged. Furthermore, he reconstructs an \u2019aleph before this lamed, though with slightly less certainty. A perusal of the published photographs and facsimiles of the fragments may seem to give Schniedewind some backing. However, as with all inscriptions, facsimiles and, in particular, photographs can be and often are misleading. Nothing may substitute for an at-hand examination of the fragments.
\nSchniedewind\u2019s proposal to read [\u2026]\u05d0\u05dc before the word divider is fraught with difficulties. First, he understands these letters to form a theophoric element of a personal name, specifically the name of the author\u2019s father. Schniedewind proposes Hazael as the author. The problem, which goes unaddressed in Schniedewind\u2019s article, is that we do not know who Hazael\u2019s father was. He is known in the Bible as a usurper. As such, his royal predecessor, Hadad-i\u1e0fri, was not his father. If Hadad-i\u1e0fri had been Hazael\u2019s father, then either Schniedewind has misread the damaged strokes at the beginning of Line 2, or the letters \u05d0\u05dc are not part of the name of Hazael\u2019s father. The other possibility is that the name of Hazael\u2019s actual father ended with the theophoric element \u05d0\u05dc\u2014. In his later article with Zuckerman, Schniedewind goes for this option, reconstructing the name as \u05d1\u05e8\u05e7\u05d0\u05dc (Baraq\u2019el). However, there are numerous other epigraphical concerns with these letters that preclude Schniedewind\u2019s suggestion\u2014not to mention historical concerns and the fact that Schniedewind and Zuckerman assume that these initial letters on Line 2 form a name. There is no reason to assume that the letters form a name, and to do so unnecessarily jeopardizes our objectivity with regard to these letters. In Schniedewind and Zuckerman\u2019s study, this is precisely the case as the identification of a lamed immediately before the first word divider on Line 2 leads them to construct the theophoric element \u05d0\u05dc\u2014. Furthermore, Schniedewind and Zuckerman\u2019s study uses only computer imaging; it does not examine the actual physical fragments themselves. Had such an examination taken place, Schniedewind and Zuckerman would have realized that lamed is not \u2018the only letter that fits the spacing, orientation and line height\u2019.<\/p>\n
Figure 3.10. Beginning of Line 2 (left) and possible restoration of lamed for damaged letter (right). The restored lamed is modelled on the lamed in Line 11 (detail of facsimile drawn by Ada Yardeni).<\/p>\n
Epigraphically, the suggestion of restoring a lamed before the word divider has immense problems. Figure 3.10 illustrates how the restoration of lamed in the position of the damaged letter leaves a considerable gap to the word divider that follows. This is quite uncharacteristic of Fragment A, and even Fragment B. This problem must cast further doubts on our previous suggestions of \u1e6deth and \u2018ayin, also.
\nFurthermore, a careful analysis of this area of the stone, from many different angles, reveals that there is scarification on the surface of the stone between the damaged letter and the word divider. This was presumably caused by breakage of the original inscription. Yet, one particular scar is deeper than the rest, prompting the likelihood that this scar is in fact the remnant of a deliberately chiselled stroke belonging to a letter.
\nWe may now posit three different possibilities regarding this area of the stone in light of the data:<\/p>\n
1. All the scarification was wholly sustained during breakage of the inscription. Thus, there is a large gap between the damaged letter we have been trying to restore and the word divider that follows.
\n2. The scarification represents a thin individual letter originally between the damaged letter we have been trying to restore to the right and the word divider to the left.
\n3. The scarification represents a deliberately chiselled stroke that belongs to the damaged letter we have been trying to restore.<\/p>\n
As we noted, the first possibility would be quite uncharacteristic within Fragment A. As such, it has little support. The second possibility is also difficult because of the nature of the deep scar in question. The scar is relatively short and protrudes down to the left from the rough line of breakage. If we are to consider this an individual letter, then we can find no letter that matches. The only possible candidate would be a taw. However, the other attestations of taw that we have in Fragment A make this a very remote possibility. From these other attestations, we see that the taw was written as a relatively long letter with its diagonal stem extending well below the writing base line. The deep scar in Line 2, however, finishes well above the writing base line. It is also on a considerably flatter angle than the stem of a taw. Furthermore, the size of a taw would impinge on the preceding damaged letter, which we have been investigating. A taw would be too wide to accommodate. The only possibility is of a very thin letter. However, the nature and placement of the deep scar in question makes the possibility of it being an individual letter very tenuous.
\nThis leaves us with the third alternative, that the deep scar forms part of the damaged letter. This means we must search for a letter that displays some curvature towards the bottom as well as an arm that extends diagonally to the left. The only letter that could fit this description is yodh. Figure 3.11 illustrates how we may indeed restore the deep scar and the damaged letter together as a yodh. Such a yodh would display slightly more curvature than most other yodhs in Fragment A. However, such a form is certainly not beyond the scope of Fragment A. Indeed, numerous yodhs attested in Fragment A do display curvature. A very comparable form may be found at the beginning of Line 10.
\nWe may make further observations about the damaged beginning of Line 2. To the right of our restored yodh, we find the definite remains of the stem of a letter (see Fig. 3.12). The rest of this letter has been obliterated in the breakage of the inscription, apart from one small engraved point which has survived to the left of it (see Fig. 3.12 for its position). The angle at which this stem lies and the small point near it means that we can narrow down the possibilities for restoring the letter. The plausible options are \u2019aleph, heh, waw, samekh, \u1e63adhe, qoph and resh. We must consider the merits of each of these letters.<\/p>\n
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).<\/p>\n
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]).<\/p>\n
Restoring this stem to an \u2019aleph is probably the least likely option. It would require the \u2019aleph to possess an inordinately long stem, far longer than any other \u2019aleph attested in Fragment A or Fragment B. It would also require us to view the engraved point as part of the horizontal stroke on the head of the \u2019aleph. However, as the point of intersection between this stroke and the stem is not extant, it must be placed at quite a high point. This means that the stem of the \u2019aleph would have to be even longer than is currently visible. Neither the \u2019alephs of Fragment A, nor those of Fragment B, have scope enough to incorporate the usage of such an irregularly shaped \u2019aleph. Thus, we may safely dismiss it as an option.
\nThe letter heh, however, fits better. It would require us to see the engraved point to the left as part of the lowest horizontal stroke of the letter. 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.
\nA more plausible suggestion is the letter samekh. As in the case of heh, 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 Fragment A indicates just how high on the stem this horizontal stroke actually appears. As such, the samekh is a very suitable candidate for restoration.
\nLess 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. 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.
\nWe 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\u2019s head. Examination 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.
\nIn order to restore this figure as a \u1e63adhe, 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 way\u2014namely, as part of another letter.
\nAt 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. Throughout the fragment we do find the tendency for two letters to be wider spaced if the second letter is a yodh. For example, the first full word of Line 2, \u05d0\u05d1\u05d9, has significantly more distance between the beth and yodh than between the \u2019aleph 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, \u05d5\u05d9\u05e9\u05db\u05d1. The first two letters of this word, waw and yodh, are more widely spaced than the other letters because of the yodh\u2019s 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 a yodh.
\nHowever, 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.
\nOur 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 \u1e63adhe. 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, \u1e25eth and \u2018ayin. 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 visible on the stone. Therefore, we cannot consider it here.
\nOne 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.
\nThe only combinations of letters which fit into this space and align with the carved strokes are \u05e1\u05d3\u05d9, \u05e6\u05d3\u05d9, \u05e1\u05d7\u05d9, \u05e6\u05d7\u05d9, \u05e1\u05e2\u05d9 and \u05e6\u05e2\u05d9. This means that the long-stemmed letter can only be a samekh or a \u1e63adhe. Physically, the letter following can be either of our three candidates (daleth, \u1e25eth or \u2018ayin). From these six combinations, three make lexical and contextual sense.<\/p>\n
These are:<\/p>\n
1. [\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af: \u2018pious acts of\u2019 or \u2018my piety\u2019,
\n2. [\u05e0]\u05e6\u05af[\u05d7]\u05d9\u05af: \u2018triumphs of\u2019 or \u2018my triumph (s)\u2019.
\n3. [\u05de]\u05e1\u05af[\u05e2]\u05d9\u05af: \u2018journeys of\u2019 or \u2018my joumey (s)\u2019.<\/p>\n
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\u2019 stems should still be visible on the stone. The line of the stone\u2019s 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 reconstructing the text to read in accordance with the first option, [\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af. Before this \u1e25eth, there is room enough for two letters or, alternatively, one letter and a word divider. Since no word dividers appear as the first character on any line of Fragment A, and since this does not occur in other monumental lapidary inscriptions, we can plausibly say that a word divider may be placed directly before the \u1e25eth of [\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af. 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 [\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af, 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 -\u05d5\u05d1 to [\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af in order to gain the reading [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af (\u2018and because of the pious acts of\u2019 or \u2018and because of my piety\u2019).
\nWe move on, then, past the first word divider to a fully preserved \u2019aleph\u2014the first letter of the word \u05d0\u05d1\u05d9. The incisions of this \u2019aleph 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 \u2019aleph 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.
\nFollowing the \u2019aleph 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 curvature starts. Since it occurs at precisely this point, the chip may be indicative 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 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\u2019s 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 \u2018head\u2019 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, head\u2014an important epigraphical datum.
\nThe third extant letter on Line 2 is a yodh. It is quite difficult to ascertain how this letter was carved, but it is clear that the top and bottom horizontal strokes were not carved as one continuous stroke along with the letter\u2019s vertical spine.
\nFollowing 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\u2019s chisel. This is telling for, since the word divider is but one small chip, it highlights the direction of the engraver\u2019s chisel. As an engraver hammers the final portion of a stroke into the stone, an impression of the chisel\u2019s 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.
\nAnother yodh marks the beginning of a new word. This yodh differs 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.
\nThe 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 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 occupies between them.
\nWhat 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 horizontal strokes. Thus, the suggestion that the engraver carved from one of the flank edges of the inscription is confirmed\u2014that is, the left edge assuming a right-handed craftsman.
\nThe 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\u2019s breakage.
\nTherefore, we may decipher Line 2 as: [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05e1\u05e7\u05af[\u2026].<\/p>\n
Line 3
\nThe 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 \u2018hook\u2019 of the waw\u2014one at the apex and another at the topmost point\u2014are consistent with our theory of the engraver\u2019s position during carving. It also suggests that the stem was the first stroke of the letter carved into the stone since the \u2018hook\u2019 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 a yodh 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.
\nAfter the fully preserved and quite angular yodh, we see a clearly legible \u0161in. Directly above the midpoint of this \u0161in is a pock mark that appears quite regular in shape. It bears resemblance to a large word divider. However, since it lies between two lines of text, it cannot be a word divider. We must, therefore, view this mark as either a natural chip in the stone perhaps sustained during smoothing of the writing face, or as a chip inflicted during breakage and reuse of the inscription.
\nAfter the \u0161in we encounter a neat kaph. Once again, we find the stem of the letter carved from bottom to top as evidenced by the shallow and deep points at the respective extremities. The middle \u2018finger\u2019 of the kaph also appears to have been carved from bottom to top, but it is difficult to ascertain the direction of carving for the left horizontal \u2018finger\u2019. Interestingly, these two \u2018fingers\u2019 do not meet the stem of the letter at one common point. This suggests that the horizontal stroke had probably been carved from left to right since the engraver would probably have started from a point common to the other \u2018finger\u2019 if he had carved in the opposite direction.
\nThis letter is followed by a very neatly carved beth. A word divider comes after this, but it has been significantly eroded. The following word begins with an \u2019aleph. The triangular area enclosed by the three strokes of this \u2019aleph has been mostly chipped away, probably through an accident of the engraver. The leftmost point of the letter is clearly the starting point for the two cross-strokes. The vertical spine of the letter repeats the method already demonstrated by other letters of carving from bottom to top. The starting point of this stroke is actually very shallow and the finishing point quite deep. As was also seen with the \u2019aleph of Line 2, this stroke was the last stroke of the letter to be carved.
\nAfter this \u2019aleph we find another regular beth which is in turn followed by a yodh. The strokes of this yodh are relatively shallow, especially at the point where the spine bends into the top horizontal stroke. This suggests that the top horizontal stroke had been carved first from right to left, after which the spine was carved as expected from bottom to top. The tail of the letter was carved from the bottom of the vertical spine out to the right.
\nWe encounter another yodh after a word divider. The oddity of this yodh is that the top horizontal stroke is abnormally long. The spine curves into both this top horizontal stroke and the tail at bottom right, giving us a precedent for the rounded yodh proposed for restoration immediately before the first word divider of Line 2. The incisions of this letter are also fairly shallow due to some erosion.
\nNext we find our first extant heh. The stem of this heh is fairly long below the bottom horizontal stroke. The lower two horizontal strokes are carved very shallow and barely manage to meet the letter stem. Beyond this heh we have a very deeply carved kaph. The left \u2018finger\u2019 of the letter has some slight chipping at its extremity. All the \u2018fingers\u2019 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 engraver\u2019s 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\u2019s 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 inscription. It would be an unnatural angle from anywhere else.
\nFollowing this word divider, only two letters are still extant. The first is an \u2019aleph 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\u2019s 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 \u2018b\u2019. The line of breakage of Fragment A impedes on this lamed at the point where the \u2018shank\u2019 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 \u2018shank\u2019 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.
\nThus, we can decipher the characters on Line 3 as: \u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05d4\u05da\u00b7\u05d0\u05dc[\u2026].<\/p>\n
Line 4
\nLine 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 \u2019aleph that follows, the vertical stroke again has been carved from bottom to top. The spine of the following \u2019aleph is also quite shallow. Very close to the leftmost point of the \u2019aleph 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 \u2018shank\u2019 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.
\nBeyond this word divider is our first fully preserved attestation of qoph. As expected, the stem of the letter has been carved in an upwards direction. Furthermore, the round head of the qoph was evidently started at the bottom left. This is to be expected as it would be the logical and most natural point to start carving such a rounded head for someone situated at the left edge of the inscription. This starting point is observable as a distinct bump or corner with shallow points. The stem seems to have been the last stroke of the letter carved and it protrudes through and above the round head.
\nAfter the qoph we find our first instance of the letter daleth. It is similar in shape to the top of a resh but characteristically it lacks the full stem. Again, vertical strokes have been carved upwards. The bottom horizontal stroke barely meets the right diagonal suggesting that this diagonal was the last stroke carved. Next to this, we see a mem, the top left portion of which is slightly eroded. Nevertheless, we can observe the trend to carve upwards in all three vertical strokes. It is, however, impossible to tell which stroke was carved first and which last. Yet, we can reason that it was carved from left to right as most other letters seem to have been.
\nFollowing the mem is a word divider made shallow by some erosion. This patch of erosion imposes on the next letter, another beth, making its strokes also shallow. After this beth we come across another \u2019aleph. As has been the case with previous instances of \u2019aleph, the leftmost point appears to have been the starting point of the letter. However, the bottom horizontal stroke appears carved from right to left because of a deep point at the left extremity. This would explain why these two strokes do not appear to merge at a common point. The spine, though, has again been carved upwards and because of its depth is obviously the final stroke of the letter.
\nA resh follows the \u2019aleph. Its incisions are quite smooth. Once more we observe the upwards direction of the vertical strokes. Next to this we have another qoph, the head of which is uniformly rounder than the one occurring earlier in the line. Once again, however, we can see that the head was carved starting from the bottom left where we observe shallow points and a bump in the roundness of the head. Unlike the earlier qoph in the line, however, this qoph\u2019s stem barely protrudes above the head. It is the last stroke of the letter to have been carved and is in fact fairly crooked.
\nA word divider occurs after this, positioned quite close to the qoph. Because of another patch of erosion, this word divider is quite shallow. After this, we have an \u2019aleph that has undergone a good deal of damage. Presumably this was caused during the inscription\u2019s breakage, as was the erosion which impinges on the preceding word divider. The enclosed triangular area of the \u2019aleph has been chipped completely out in a way similar to that of the first \u2019aleph in Line 3. However, the leftmost portion of this \u2019aleph has been totally damaged because of a very deep lacuna in the stone. There is no difficulty, though, in identifying this letter as an \u2019aleph.
\nThis lacuna presents an interesting problem that received only cursory attention from Biran and Naveh, and none from any other scholar concerned with the fragment. The problem is the size of the lacuna and the gap between the \u2019aleph just mentioned and the beth which emerges clearly (albeit slightly damaged) at the end of the lacuna. Biran and Naveh sufficed to say that this gap did not warrant the restoration of a letter in between the \u2019aleph and the beth. Rather, they suggested that a small blemish already existed on the stone prior to the engraving of the inscription and it was this which accounted for the larger than usual gap. This blemish, they claim, then became larger over time.
\nHowever, 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 certainly 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 \u2019aleph and the beth should be so far apart.
\nIn 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 \u2019aleph 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 misleading. The lighting in this photograph, taken by Radovan, gives the impression 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 photograph. The \u2018scar\u2019 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 contrasted image of the fragment shows the depths of the scar and deliberately carved stroke a little more clearly (see Fig. 3.13).<\/p>\n
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; photograph: Z. Radovan [detail]; facsimile drawn by Ada Yardeni [detail]).<\/p>\n
Having observed this carved stroke, we must now identify it with a letter 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, we would need to posit a taw with a significant downward bend in the stem for it to even fit. The slant of the stroke protruding from the lacuna means that a straight stemmed taw would run into the next letter, beth. Yet, such a form of taw with the stem bent downwards is not attested among the fragments. We find forms in which the stem curves upwards, but this would pose greater problems here. As such, we must reject taw as a possible restoration.
\nThe lamed, however, fits perfectly. The visible remnant forms part of the \u2018shank\u2019 of the letter which would have curved underneath itself to the right in order to form the hook (see Fig. 3.14). This accords perfectly in shape and position with all other lameds attested in both Fragment A and Fragment B. Although this gives the three letters over which the lacuna impinges a slightly crammed appearance, it is nothing extraordinary to Fragment A. On the contrary, we observe many similar examples of clustered letters. Most notably, from Line 6, the word \u05de\u05dc\u05db\u05d9 provides an excellent example of a lamed being squashed close to the preceding letter. Therefore, lamed is the letter that we must restore here.
\nLine 4 should, therefore, be deciphered as: \u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd\u00b7\u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05dc\u05af\u05d1\u05d9[\u2026].<\/p>\n
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]).<\/p>\n
Line 5
\nThe fifth line of Fragment A begins with a smoothly eroded \u2019aleph. This erosion is part of the belt of erosion that covers the right edge of the entire fragment. Following the \u2019aleph, we come to our first attestation of the letter nun. Because of the erosion along the inscription\u2019s 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\u2019s horizontal stroke is harder to discern.
\nThe nun is followed by a heh, the stem of which is fairly short. As we have seen with other occurrences of heh, the stem was the final stroke of the letter to be carved, and it was carved upwards. A word divider follows very closely, almost touching the top \u2018bristle\u2019 of the heh. It is quite small and shallow due to the smooth belt of erosion.
\nAfter this comes a waw that is also quite shallow due to erosion. However, 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 upwards 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 \u2018bristles\u2019 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 upwards and the horizontal \u2018finger\u2019 was carved leftwards. The top of the stem also comes very close to touching the top \u2018bristle\u2019 of the preceding heh.
\nA 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\u2019s tool in the upper right part of the indentation. Again, the engraver was evidently situated at the left flank of the inscription.
\nAnother 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.
\nTwo 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 sustained during or after breakage of the inscription. The bottom horizontals of these two daleths are different from each other in that the first slants slightly upward on the right whereas the second slants downwards. On both forms, however, the two upright strokes forming the \u2018roof\u2019 of the daleth have been carved from the bottom up. Because of the chip along the bottom horizontal of the first daleth, it is difficult to see in which direction this stroke was carved. However, we can discern from the second daleth that its horizontal was carved from left to right.
\nAnother shallow word divider follows the two daleths. Again, we can observe clearly the direction of carving, suggesting the engraver\u2019s position at the left flank of the inscription. A fully preserved qoph follows this. As has been the case with previous attestations of qoph, we can observe the point at which the engraver started carving the rounded head of the letter in the bottom left portion. Again, this is distinguishable by a slight bump in the roundedness of the head. This is followed by another neat daleth that displays similar characteristics to the first daleth encountered earlier in the line. A mem follows that, which is quite neat and regular. We observe once more the upward direction of carving in each vertical stroke. The final extant letter of the line is undoubtedly a yodh that has been damaged by some chips in the stone in the lower portions of the letter. The interesting feature of this yodh is that it is quite curved when compared with some of the other more angular forms in Fragment A. It closely resembles the Arabic numeral \u20182\u2019 with an extra horizontal in the middle. From the shape of the chipping over the letter, it appears that the engraved strokes facilitated the damage.
\nBeyond this is the leftmost extant point of the written surface on Fragment A. At the extremity are the minute remains of another character. However, this is virtually impossible to decipher on its own because there is hardly anything to work with. We may, however, say that since the preceding letters form a coherent word, \u05e7\u05d3\u05de\u05d9, a word divider is to be expected. Indeed, the small inscribing is in the perfect position for a word divider. As such, I suggest restoring this small carving to a word divider. We may, then, decipher Line 5 as: \u05d0\u05e0\u05d4\u00b7\u05d5\u05d9\u05d4\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05e7\u05d3\u05de\u05d9\u00b7\u05af[\u2026].<\/p>\n
Line 6
\nThe first letter of Line 6 is a yodh 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 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.
\nThe 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 \u2018finger\u2019 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\u2019s position during carving.
\nA 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 \u2019aleph follows after this, which is very regular and displays all the normal characteristics of previous \u2019alephs. There has been some slight abrasion of the stone across the bottom of the letter.
\nNext 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 \u2018shank\u2019 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.
\nFollowing 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.
\nAfter 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 semicircular 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 \u0161in.
\nThe 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. 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. However, 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 theoretically possible but highly unlikely.
\nWe 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 \u2019aleph in that line posed no serious problem for restoring a lamed because of the nature of the lamed\u2019s 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 announced the discovery of the B fragments, Biran and Naveh changed their opinion to also restore this letter as a lamed. 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.
\nUnfortunately, the recess covers over most of the remaining letters of the line on the fragment. However, the edge of the recess is in such a position that if any of the letters it covers were a lamed, we would still be able to see the top portion of the \u2018shank\u2019. We do not find any such strokes beyond the one just dealt with, so there must be no other lamed on the rest of the extant line.
\nWe do, however, still find the remnant of a character at the very end of this line as it is extant on Fragment A. This letter remnant is situated along the broken edge precisely at the point where the recess breaks off completely with the edge of the fragment. The remnant consists of three strokes that closely resemble the top of a waw. Unfortunately, the photograph by Radovan does not highlight these strokes well enough. Also, the facsimile by Yardeni does not show the strokes in their entirety. A thorough examination of the actual fragment, however, shows these three strokes quite distinctly (see Fig. 3.15).<\/p>\n
Figure 3.15. Detail of the facsimile 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.<\/p>\n
We cannot find any other letter fitting this shape, besides waw. Yardeni suggested that this was the tip of a nun. However, this has only taken into account the rightmost of the three strokes\u2014a 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 \u2018hook\u2019. Dijkstra, who also noticed only one stroke, suggested that it was the remnant of a \u0161in. He used images of the fragment to transpose letters over this stroke and concluded that \u2018the small stroke almost exactly coincided with the rightside of the letter, sin of y\u015br\u2019l\u2019. However, this is simply not the 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 \u2019aleph. In the facsimile published with his article, we observe that Puech noticed more than one stroke at the edge of the fragment here. However, he misconstrued the precise nature and orientation of these strokes, and so came to a faulty conclusion.
\nThe only letter we may restore here is a waw. Indeed, we find almost identical forms of waw elsewhere among the fragments. The waw occurring earlier in Line 6 shows a form where the stem displays a notable kink at the junction with the \u2018hook\u2019. The letter remnant at the edge of the fragment 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 \u2018hook\u2019, 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.
\nThis 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 letters 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 word divider poses little difficulty in terms of spacing or philology. Thus, we suggest the restoration of one letter and a word divider in this gap. In the context of the inscription, to be discussed in a later chapter, I contend that we should restore a kaph and a word divider here. This fits not just the written context, but the physical epigraphic evidence, also.
\nWe can restore Line 6, then, as: \u05d9\u00b7\u05de\u05dc\u05db\u05d9\u00b7\u05d5\u05d0\u05e7\u05ea\u05dc\u00b7\u05de\u05af\u05dc\u05af[\u05da\u00b7]\u05d5\u05af[\u2026].<\/p>\n
Line 7
\nDue 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\u2019s 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 \u2019aleph follows this waw which has significantly sharper edges and contours 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.
\nOur 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.
\nA 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 the yodh restored immediately before the first word divider on Line 2.
\nAfter 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 indicating it was carved with two strokes meeting at one point.
\nFollowing 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. A regular \u0161in 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 suggest any restoration for this letter based solely on the physical evidence.
\nWe can, therefore, decipher Line 7 as: \u05db\u05d1\u00b7\u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9\u00b7[\u2026].<\/p>\n
Line 8
\nLine 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.
\nAfter 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 the yodh that follows. That is, it imposes upon the space of the yodh. 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.
\nThe 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 \u05d9\u05e9\u05e8\u05d0\u05dc, of which all the letters appear regular.
\nAfter the lamed of \u05d9\u05e9\u05e8\u05d0\u05dc 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 waw. After this waw we have half of a qoph. The edge of the large recess follows precisely down the length of the qoph\u2019s 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 of qoph.
\nAfter 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 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. 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: \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc\u00b7\u05d5\u05e7\u05af\u05ea\u05af\u05dc\u05af[\u2026].<\/p>\n
Line 9
\nThe first letter of Line 9 is a kaph that shows signs of wear due to continued erosion. The \u2018fingers\u2019 of the kaph show the most erosion. Interestingly, the \u2018fingers\u2019 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.
\nAfter a shallow word divider, we encounter a beth which is smaller in size than other beths in Fragment A. The strokes which form the \u2018nose\u2019 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.
\nFollowing 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 \u2018hook\u2019 of this waw appears at an odd slant, lending it a \u2018heavy\u2019 appearance. An \u2019aleph follows this, which possesses a rather long stem.
\nAfter this \u2019aleph, 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 \u0161in. A very close examination of the point of breakage 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 restoration of a yodh, 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 consider one at all. A \u0161in, however, has no such problems and is therefore to be preferred.
\nBeyond 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 \u0161in. 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 \u0161in. However, the preceding comparison of these two letters in Line 1 showed that the leftmost stroke of all \u0161ins 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 \u0161in here because the evidence from the available sample of fragments is unanimous in this difference between \u0161in and mem. After this restored mem, we see a word divider. At this point, the line breaks off.
\nWe can decipher Line 9, therefore, as: \u05da\u00b7\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3\u00b7\u05d5\u05d0\u05e9\u05af\u05dd\u05af\u00b7[\u2026].<\/p>\n
Line 10
\nThe beginning of Line 10 has a shallow yodh, followed by a taw and a very eroded word divider. This is followed by an \u2019aleph, 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 \u2018nose\u2019 of this resh, affecting the following letter, a qoph. Interestingly, the head of this qoph has a double crown unlike other attestations. Another shallow word divider appears after this.
\nNext 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. Unfortunately, the line of breakage occurs before the next letter.
\nLine 10, then, should be deciphered as: \u05d9\u05ea\u00b7\u05d0\u05e8\u05e7\u00b7\u05d4\u05dd\u00b7\u05dc[\u2026].<\/p>\n
Line 11
\nA shallow \u2019aleph begins Line 11. After it we encounter our first instance of the letter \u1e25eth. The bottom \u2018rung\u2019 of the \u1e25eth has been damaged by erosion, but it otherwise appears to have been carved very neatly. The vertical \u2018runners\u2019 of the letter have been carved from the bottom up while the top two \u2018rungs\u2019 have been carved from left to right. Due to erosion, it is impossible to tell in which direction the bottom \u2018rung\u2019 was carved. We may notice, though, that the middle \u2018rung\u2019 is shallow in comparison with the other strokes, suggesting that it was the last \u2018rung\u2019 carved with extra caution.
\nAfter this \u1e25eth 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 \u2018hook\u2019 of this waw is fairly round rather than angular. The lamed that follows, though, is actually 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 incisions do not permit us to identify the letter more accurately than this.
\nWe can decipher Line 11, then, as: \u05d0\u05d7\u05e8\u05df\u00b7\u05d5\u05dc\u05d4[?\u2026].<\/p>\n
Line 12
\nVery 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 incision 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.
\nThe 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, particularly around the \u2018fingers\u2019 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.
\nAfter a very shallow word divider, we encounter a regular \u2018ayin. 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 \u0161in. Only the bottom left portion of the \u0161in has been broken off so its identification is not in doubt.
\nWe may decipher the extant characters on Line 12, therefore, as: \u05dc\u05da\u00b7\u05e2\u05dc\u00b7\u05d9\u05e9[\u2026]. Reconstruction of some of the non-extant letters will be left for the textual analysis.<\/p>\n
Line 13
\nThe 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 breakage 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 \u1e63adhe. The stem of this \u1e63adhe 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 certainty the right-hand side of the letter. Its identification as a \u1e63adhe, though, is not in doubt since no other letter in Fragment A compares with it, and the \u1e63adhe of other inscriptions is comparable with what is still visible.
\nThe next letter is a very upright resh. Between it and the preceding \u1e63adhe, 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 \u2018ayin. 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.
\nLine 13 can, therefore, be deciphered as: \u05de\u05e6\u05af\u05e8\u00b7\u05e2\u05dc[\u2026].<\/p>\n
Preserved Side
\nSome 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\u2019s 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 concentrated near the edge, as it is in the written surface, it was most probably sustained during its secondary usage as building material. We would expect a broader spread of smooth erosion over both the written surface and this preserved face if the erosion was sustained while the inscription was in primary display.
\nThere is no sign of any writing or pictographic carving of any kind on this preserved side.<\/p>\n
Newly Exposed Surfaces and Claims of Forgery
\nDue to the stone\u2019s breakage, many new faces of the stone have been exposed. In some places, chipping of these exposed faces has occurred. The most notable appears in the surface exposed at the top of the fragment and is clearly visible from the side of the fragment. These chips were understood by Cryer to denote deliberate chisel marks. As a result, Cryer denounced the fragments as a forgery at a postgraduate seminar for students of Copenhagen University. His reasoning was that if the fragments were genuine, there would have been no reason to use a chisel on the newly exposed surface after its breakage and during its transference to its secondary position in a wall at the outer edge of the piazza. Therefore, a chisel mark could only have been made during the carving of the inscription and so the stone must have been broken prior to the carving of the text. Inevitably, this would mean that the fragment could not claim to be part of an original whole inscription, thereby rendering it a forgery.
\nCryer came to this conclusion of questionable authenticity after having published four articles on the Tel Dan fragments. Prior to his claim of forgery, Cryer had advocated seeing the fragments as two separate inscriptions. His opinion was swayed after seeing the fragments on display at the Israel Museum, Jerusalem. According to Cryer, the nature of these chips as chisel marks was confirmed by an independent engraver who looked over them.
\nUnfortunately for Cryer, there is nothing about the chips in question which requires us to see any of them as the result of chisel blows. Very close examination of these chips on the exposed surfaces demonstrates that there are no telling traces of a chisel at their deepest point. This would be expected if, as Cryer claims, the chips had been made by a chisel. Rather, most appear to be naturally made chips, while the deepest of them had probably been created by collision with some type of jagged edge, probably that of another rock. Thus, we expect them to have been sustained when the stone was destroyed or, more likely, when it was repositioned in the wall at the perimeter of the piazza.
\nIn addition to the physical evidence, it is hard to imagine why a potential 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 categorically 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.
\nFurthermore, 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 erosion 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 authenticity of Fragment A.
\nCryer\u2019s 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, palaeographical, philological and historical expertise.
\nNeither 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 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\u2019s 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\u201393 with such unscrupulous behaviour.
\nCryer advocated that all three fragments must be seen as part of the same forgery. Indeed, if Fragments B1 and B2 represent a separate inscription, Cryer\u2019s accusations fall through. Yet, Fragments B1 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 B1 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 summarily dismissed.<\/p>\n
Synthesis of the Script of Fragment A and the Script of Fragment B<\/p>\n
The date arrived at for the script of Fragment B (c. 800 BCE, +\/\u2212 20 years) is confirmed by the fact that the script of Fragment B is very close, if not identical, to the script of Fragment A. If we perform a statistical analysis of the relationship between the fragments for each individual letter, we see this fact substantiated. First, four letters (gimel, zayin, teth and \u1e63adhe) are either attested in only one fragment or in neither. Therefore, we may dismiss them for purposes of considering a relationship between the fragments. Of the remaining 18 letters, ten (56 per cent) show a very close correspondence and six (33 per cent) appear to be closely related variants. The remaining two (11 per cent) letters (nun and taw) are special cases in that they are incomplete figures in Fragment B and difficult to compare with their counterparts in Fragment A. Most importantly, no letters are irreconcilably different.
\nThis last particular point may legitimately be claimed for many contemporary but mutually exclusive inscriptions. As such, the fact that no letters are irreconcilably different is no guarantee that the fragments are to be understood as belonging to the same inscription. However, it does much to undermine the objections of Cryer, Becking and others who claim that the difference between the fragments demand the interpretation that the fragments are mutually exclusive. This is certainly not the case.
\nFurthermore, two particular letters are invaluable in actually clinching a relationship between Fragment A and Fragment B. First, the letter waw is noted by both Cryer and Becking as probably the most obvious and telling figure for the mutual exclusivity of the fragments. However, close observation of the forms in both fragments reveals that the forms do in fact overlap quite significantly. As such, the letter waw provides a bridge rather than a chasm between Fragment A and Fragment B.
\nSecond, 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 \u2018jagged head\u2019 had slipped slightly downwards. This internal inconsistency alerts us to the tendency of the scribe of Fragment B to make slight alterations 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 \u2018bristles\u2019 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 established. 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 palaeographical analysis also confirms the veracity of the date for the fragments yielded in our examination of the archaeological context. As such, we may legitimately claim that the Tel Dan Inscription was written at some time close to 800 BCE.<\/p>\n
Summary of Script Analysis<\/p>\n
Tables 4.40 and 4.41 summarize the palaeographical information gleaned on each set of letters from both Fragment A and Fragment B.<\/p>\n
Table 4.40. Summary of Palaeographical Information for the Script of Fragment A.<\/p>\n
FRAGMENT A
\nLetter
\nClosest Matching Inscriptions
\nRegional Bracket
\nChronological Bracket (centuries BCE)
\n\u05d0
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nHamath Bricks
\nNerab Stelae
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia Syria
\nTransjordan
\nLate ninth\u2013early fifth
\n\u05d1
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nEqron
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nTema
\nNorthwest Arabia
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate ninth\u2013mid-fifth
\n\u05d2
\n\u2014
\n\u2014
\n\u2014
\n\u05d3
\nKilamuwa
\nAmman Citadel
\n(Samaria Ostraca)
\nHamath Bricks
\nBar Rakib
\nSiloam Tunnel
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate-ninth\u2013late eighth
\n\u05d4
\nMelqart Stele
\nMesha Stele
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nHazael\u2019s Horse Ornament
\nZakkur
\nBar Rakib
\nNerab Stelae
\nEqron
\nCyprus
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05d5
\nKilamuwa
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nBar Rakib
\nCyprus
\nSoutheast Anatolia
\nLate ninth\u2013late-eighth
\n\u05d6
\n\u2014
\n\u2014
\n\u2014
\n\u05d7
\nTell Fakhariyah
\nHazael Ivory
\nKilamuwa
\nAmman Citadel
\nZakkur
\nSiloam Tunnel
\nPalestine
\nSamal
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05d8
\n\u2014
\n\u2014
\n\u2014
\n\u05d9
\nMelqart Stele
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nAzitawadda
\nBar Rakib
\nNimrud Ivory Plaque
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05db
\nMesha Stele
\nKuntilet \u2018Ajrud
\nAmman Citadel
\nNora Inscription
\nPhoenician Governor\u2019s Bowl
\nPanammu I
\nHamath Bricks
\nBar Rakib
\nSiloam Tunnel
\nEqron
\nCyprus
\nNegev
\nPalestine
\nSE Anatolia
\nSardinia
\nSyria
\nTransjordan
\nLate ninth\u2013early seventh
\n\u05dc
\nA\u1e25iram Sarcophagus
\nYe\u1e25imilk shipi\u1e6dba\u02bfal
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nKuntilet \u2018Ajrud Wares
\nZakkur
\n(Samaria Ostraca)
\n(Samaria Seals)
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\n(Khirbet el-Qom)
\n(Ophel Ostracon)
\nAzitawadda
\nPanammu II
\nBar Rakib
\n(Tel Qasil\u00e9 Ostracon II)
\n(Beersheba Juglet)
\nNerab Stelae
\n(Lachish III Seals)
\n(Ramat Ra\u1e25el Seal)
\n(Tell en-Na\u1e63beh Seal)
\n(Wadi Muraba\u2018at Papyrus)
\n(Shechem III Seal)
\nTema
\n(Yesha\u2018yahu Bulla)
\nCyprus
\nNegev
\nNorthwest Arabia
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-tenth\u2013mid-fourth
\n\u05de
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nBar Rakib
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05e0
\nMelqart Stele
\nMesha Stele
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nSa\u1e43al
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05e1
\nKilamuwa Scepter
\nAmman Citadel
\nHamath Bricks
\nAzitawadda
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate ninth\u2013late eighth
\n\u05e2
\nMesha Stele
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nPanammu II
\nNerab Stelae
\nSiloam Tunnel
\nSilwan Tomb
\nEqron
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early fifth
\n\u05e4
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nLate ninth\u2013early fifth
\n\u05e6
\nComparison not practicable
\n\u2014
\n\u2014
\n\u05e7
\nMelqart Stele
\nHazael\u2019s Horse Ornament
\nKilamuwa
\nKilamuwa\u2019s Sceptre
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nMid-ninth\u2013early seventh
\n\u05e8
\nMelqart Stele
\nMesha Stele
\nKilamuwa
\nKilamuwa Sceptre
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nTema
\nCyprus
\nNorthwest Arabia
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013mid-fifth
\n\u05e9
\nTell Fakhariyah
\nAmman Citadel
\nNora
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nNerab Stelae
\nEqron
\nCyprus
\nMesopotamia
\nPalestine
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05ea
\nMelqart Stele
\nHazael Ivory
\nHazael\u2019s Horse Ornament
\nArchaic Phoenician (Cyprus)
\nKilamuwa
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nAzitawadda
\nPanammu II
\nBar Rakib
\nNerab Stelae
\nCyprus
\nSoutheast Anatolia
\nSyria
\nMid-ninth\u2013late eighth
\nDate for script of Fragment A: c. 800 BCE (\u00b1 20 years)<\/p>\n
Table 4.41. Summary of Palaeographical Information for the Script of Fragment B.<\/p>\n
FRAGMENT B
\nLetter
\nClosest Matching Inscriptions
\nRegional Bracket
\nChronological Bracket (centuries BCE)
\n\u05d0
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nHamath Bricks
\nNerab Stelae
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nLate ninth\u2013early fifth
\n\u05d1
\nMelqart Stele
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nBar Rakib
\nNerab Stelae
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\n(Tema)
\nCyprus
\n(Northwest Arabia)
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nLate ninth\u2013early fifth
\n\u05d2
\nYe\u1e25imilk
\nEliaba\u2018al
\nKilamuwa
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nPanammu I
\nPanammu II
\nSefir\u00e9 Treaties
\nBar Rakib
\nNerab Stelae
\nTabnit Sarcophagus
\nTema
\nCyprus
\nNorthwest Arabia
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nLate tenth\u2013mid-fifth
\n\u05d3
\nKilamuwa
\nAmman Citadel
\n(Samaria Ostraca)
\nHamath Bricks
\nBar Rakib
\nSiloam Tunnel
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate-ninth\u2013late eighth
\n\u05d4
\nMelqart Stele
\nHazael\u2019s Horse Ornament
\nKilamuwa
\nAmman Citadel
\nNora Inscription
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nBar Rakib
\nSilwan Tomb
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05d5
\nKilamuwa
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nBar Rakib
\nCyprus
\nSoutheast Anatolia
\nSyria
\nLate ninth\u2013late eighth
\n\u05d6
\nMelqart Stele
\nKilamuwa
\nKilamuwa\u2019s Sceptre
\nPanammu I
\nAzitawadda
\nSefir\u00e9 Treaties
\nSiloam Tunnel
\nSilwan Tomb
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05d7
\nTell Fakhariyah
\nHazael Ivory
\nKilamuwa
\nAmman Citadel
\nZakkur
\nSiloam Tunnel
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05d8
\n\u2014
\n\u2014
\n\u2014
\n\u05d9
\nMelqart Stele
\nKilamuwa
\nAmman Citadel Inscription
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nBar Rakib
\nAzitawadda
\nNimrud Ivory Plaque
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05db
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel Inscription
\nKuntilet \u2018Ajrud
\nNimrud Ivory Plaque
\nAzitawadda
\nSiloam Tunnel
\nEqron
\nNegev
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05dc
\nA\u1e25iram Sarcophagus
\nYe\u1e25imilk
\nShipi\u1e6dba\u02bfal
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nArchaic Cyprus
\nNora Inscription
\nKuntilet \u2018Ajrud Wares
\nZakkur
\n(Samaria Ostraca)
\n(Samaria Seals)
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\n(Khirbet el-Qom)
\n(Ophel Ostracon)
\nAzitawadda
\nPanammu II
\nBar Rakib
\n(Tel Qasil\u00e9 Ostracon II)
\n(Beersheba Juglet)
\nNerab Stelae
\n(Lachish III Seals)
\n(Ramat Ra\u1e25el Seal)
\n(Tell en-Na\u1e63beh Seal)
\n(Wadi Muraba\u2018at Papyrus)
\n(Shechem III Seal)
\nTema
\nYesha\u2018yahu Bulla
\nCyprus
\nNegev
\nNorthwest Arabia
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-tenth\u2013mid-fourth
\n\u05de
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nBar Rakib
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05e0
\nComparison not practicable
\n\u2014
\n\u2014
\n\u05e1
\nKilamuwa Sceptre
\nAmman Citadel
\nHamath Bricks
\nAzitawadda
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate ninth\u2013late eighth
\n\u05e2
\nMesha Stele
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nPanammu II
\nNerab Stelae
\nSiloam Tunnel
\nSilwan Tomb
\nEqron
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSyria
\nTransjorden
\nMid-ninth\u2013early fifth
\n\u05e4
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nLate ninth\u2013early fifth
\n\u05e6
\n\u2014
\n\u2014
\n\u2014
\n\u05e7
\nMelqart Stele
\nHazael\u2019s Horse Ornament
\nKilamuwa
\nKilamuwa Sceptre
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nMid-ninth\u2013early seventh
\n\u05e8
\nMelqart Stele
\nMesha Stele
\nKilamuwa
\nKilamuwa Sceptre
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSyria
\nTransjorden
\nMid-ninth\u2013mid-fifth
\n\u05e9
\nTell Fakhariyah
\nAmman Citadel
\nNora Inscription
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nNerab Stelae
\nEqron
\nCyprus
\nMesopotamia
\nPalestine
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-ninth-early seventh
\n\u05ea
\nComparison not practicable
\nSyrian Influence
\n\u2014
\nDate for script of Fragment B: c. 800 BCE (\u00b1 20 years)<\/p>\n
Fragment B<\/p>\n
A facsimile of Fragment B (B1 and B2 joined together) was drawn by Ada Yardeni (Fig. 4.2).<\/p>\n
Figure 4.2. Facsimile of Fragment B (B1+ B2; drawn by Ada Yardeni).<\/p>\n
Analysis of Individual Letters
\n\u2019aleph. There are five instances of the letter \u2019aleph in Fragment B. Table 4.20 contains their images in sequential order.<\/p>\n
Table 4.20. Occurrences of \u2019aleph in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 2
\n\u05d1\u05d0[?\u2026]
\nB-2
\nLine 4
\n\u05d0\u05af[\u2026]
\nB-3
\nLine 5
\n[\u2026]\u05d0\u05e4\u05e7
\nB-4
\nLine 6
\n\u05d0\u05af\u05e1\u05e8\u05d9
\nB-5
\nLine 6
\n\u05d0[\u2026]<\/p>\n
The form of the figure is very similar, if not identical, to the form of \u2019aleph observed in Fragment A. The forms are particularly close to the \u2019alephs occurring in the latter portions of Fragment A. In particular, \u2019alephs 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.
\nOf particular interest is the tendency to kink the top of the stem somewhat leftward. This idiosyncrasy was noticed in Fragment A also. Thus, it is quite reasonable to suggest that the same hand lay behind both fragments. Yet, caution is needed for we see the same peculiar kink in other inscriptions, most notably the Amman Citadel Inscription and the Sefir\u00e9 Inscriptions. Thus, the slight kink in the stem may not be characteristic of an individual\u2019s writing technique, but of a prevailing style. Since it is such a small detail, however, it is difficult to determine which is the case.
\nSince there is such a close match between the forms of \u2019aleph 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 Sefir\u00e9 Treaties, the Hamath Bricks, the Nerab Stelae and the sarcophagi of Tabnit and Eshmun\u2018azar.
\nThus, the \u2019aleph 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.<\/p>\n
Beth. There are four instances of the letter beth in Fragment B. Their images are shown in Table 4.21 in sequential order.<\/p>\n
Table 4.21. Occurrences of beth in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccuring Word
\nB-1
\nLine 2
\n\u05d1\u05d0[?\u2026]
\nB-2
\nLine 5
\n\u05e9\u05af\u05d1\u05e2\u05af[\u2026]
\nB-3
\nLine 7
\n\u05d1\u05e8
\nB-4
\nLine 8
\n\u05d1\u05e8[\u2026]<\/p>\n
In each case, the figure is notably rotated in an anti-clockwise direction. The sole example in Fragment B1, however, possesses a much larger head in proportion to the examples occurring in Fragment B2. The point at which the stem of the figure bends beneath the head to form a tail is quite distinct. Only in beth B-4 is this vertex somewhat softened, hinting towards a style of curved-stem beth. The tail is also fairly straight.
\nIn these latter two features of a vertexed-stem form, Fragment B differs considerably from Fragment A, which employs curved-stem beths. It seems, at first, an irreconcilable difference of style. However, we did note in the analysis of Fragment A the concurrency of both the vertexed-stem and curved-stem beths, even within the same inscription. This was adequately demonstrated by the Sefir\u00e9 Treaties, in particular Stelae I and III. To this we may add that both styles appear side by side in the Amman Citadel Inscription and the Eqron Inscription. Line 2 of the Amman Citadel Inscription contains two beths standing side by side. The first of these beths corresponds in form to the curved-stem beths common in Fragment A. The second, however, corresponds to the vertexed-stem beths of Fragment B. Indeed, throughout the eight extant lines of the Amman Citadel inscription, these two types of beth coexist. Turning to the Eqron Inscription, we find both styles of beth used in the first line. Therefore, the seemingly different styles of beth employed in Fragment A and Fragment B are far from evidence of a different hand. Indeed, the curved-stem beth is not attested without the vertexed-stem beth, except in tenth century BCE Phoenicia\u2014a milieu that does not fit either Fragment A or Fragment B. That is, the curved-stem beth is not mutually exclusive from the vertexed-stem beth at this time. Thus, the alternate styles in the Tel Dan fragments complement each other and actually support the placing of both Iragments together under the one hand.
\nIn addition, we can make an appeal to beths A-1 and A-5 in Fragment A, which hint at the vertexed-stem form of beth. Thus, there is ample evidence to suggest a bridge between the styles of Fragment A and Fragment B with regards to the letter beth.
\nRemembering that the vertexed-stem beth is the most common style of beth used in Northwest Semitic inscriptions outside tenth century BCE Phoenicia, we see Fragment B indicating an affinity with the popular style. Coupled with the all-important slant of the whole figure, we may date the beth of Fragment B between the mid-ninth century BCE and the early fifth century BCE. The Melqart Stele is the earliest attestation of a comparable form, and the latest are the sarcophagi of Tabnit and Eshmun\u2018azar, and possibly the Tema Stele. Between these dates, we see particular affinity with Kilamuwa, Nora, the Amman Citadel Inscription, the Archaic Phoenician Inscription from Cyprus, Panammu I and the Sefir\u00e9 Treaties. The Zakkur Stele has comparable forms except for a rounded \u2018nose\u2019. Panammu II, Bar Rakib and the Nerab Stelae represent a shift in style to a more upright letter, but nonetheless yield some good examples. Thus, the latter inscriptions of Eshmun\u2018azar 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.<\/p>\n
Gimel. There is a single instance of the letter gimel in Fragment B. It occurs in the first line of Fragment B1 and is displayed in Table 4.22.<\/p>\n
Table 4.22. Occurrences of gimel in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 1
\n[\u2026]\u05d5\u05d2\u05d6\u05e8\u05af[\u2026]<\/p>\n
The figure consists of two straight strokes, which meet at a common vertex, much like a figure \u20187\u2019. However, the entire letter is rotated significantly anti-clockwise so that the vertex forms the highest point of the letter. The right stroke is discemibly longer than the left stroke and, therefore, forms what we may label as the stem. To be precise, the left stroke is two-thirds the length of the stem and the angle formed at the vertex is 58\u00b0.
\nSince no gimel is attested in Fragment A, we can make no comparison with the form here. Thus, the letter gimel is inconsequential in our attempt to determine a relationship between Fragment A and Fragment B. Although there is only a single example here in Fragment B, the simplicity of the figure means we can be fairly sure that it represents a normal form employed by the scribe of Fragment B and attempt a dating of it.
\nWe may go back as far as the Proto-Canaanite of the thirteenth century BCE to find a match. However, since the script of Fragment B shows a good deal of development from Proto-Canaanite, we cannot legitimately suggest it as a match. Rather, the similarities are due to the simple nature of the letter rather than to any synchronism.
\nThe gimels attested in tenth century BCE Byblian inscriptions are the earliest meaningful matches. The inscriptions of Yehimilk and Eliba\u2019al both display the same gimel with correct proportions and rotation. After this time, good matches can be made with Kilamuwa, the Archaic Phoenician Inscription from Cyprus, Nora, Panammu I, Panammu II and Bar Rakib. Unfortunately, gimel is unattested in the Melqart Stele and the Amman Citadel Inscription. The Mesha Stele and later inscriptions from Palestine 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 northern influence. The Sefir\u00e9 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\u2019s 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.
\nThus, 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 B1 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.<\/p>\n
Daleth. There are two instances of the letter daleth in Fragment B. Both occur next to each other in the fourth line of Fragment B1. Table 4.23 contains both their images.<\/p>\n
Table 4.23. Occurrences of daleth in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 4
\n\u05d4\u05d3\u05d3
\nB-2
\nLine 4
\n\u05d4\u05d3\u05d3<\/p>\n
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 \u2019aleph. However, such is not the case. This \u2018tail\u2019-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.
\nThere 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 comparison between daleth A-2 and daleth B-2 reveals how close their respective forms are. A survey of the daleths in the Amman Citadel Inscription also demonstrates the use of similarly divergent forms.
\nAnother 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 Kilamuwa Inscription and the Amman Citadel Inscription. The only other distinguished matches are the bricks from Hamath, the stele of Bar Rakib, the Siloam Tunnel Inscription, and even the ceramic texts of the Samaria Ostraca.
\nWith 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.<\/p>\n
Heh. There are five instances of the letter heh in Fragments B1 and B2. Table 4.24 contains their images in order of their appearance.
\nThe form consists of a leftward leaning stem from which three \u2018bristles\u2019 protrude to the left. These \u2018bristles\u2019 hang down heavily so that there is an acute angle between the \u2018bristles\u2019 and the stem. The \u2018bristles\u2019 also appear cluttered very close to each other. From the three fully preserved figures we observe that the bottom \u2018bristle\u2019 branches from the approximate midpoint of the stem. Likewise, on these three particular figures we observe that the middle \u2018bristle\u2019 is shorter than the other two.<\/p>\n
Table 4.24. Occurrences of he in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 2
\n[\u2026]\u05ea\u05af\u05dc\u05d7\u05de\u05d4
\nB-2
\nLine 3
\n[\u2026]\u05d4\u05af
\nB-3
\nLine 4
\n\u05d4\u05de\u05dc\u05da
\nB-4
\nLine 4
\n\u05d4\u05d3\u05d3
\nB-5
\nLine 8
\n[\u2026]\u05d9\u05af\u05d4\u05d5<\/p>\n
Having noted the characteristics of the form on Fragment B, we see that there are some differences with the form in Fragment A. First, in Fragment A, the \u2018bristles\u2019 do not appear as cluttered together as they do in Fragment B. Also, there is a tendency in Fragment A for the \u2018bristles\u2019 to become shorter further down the stem. This is by no means universal in Fragment A, but it is a noted difference. Thus, the distinction between the two forms in this regard is the length of the bottom \u2018bristle\u2019. Also, the form of Fragment A appears to have a shorter stem than the form of Fragment B.
\nWith these divergent characteristics in mind, we must ask whether they represent irreconcilable differences between the two fragments. We must deal with each point in turn.
\nFirst, the relative clutter of the \u2018bristles\u2019 in Fragment B by no means requires the interpretation that the fragments are mutually exclusive. Such cluttering of the \u2018bristles\u2019 is often seen in other Northwest Semitic inscriptions together with less cluttered figures, the type of which are seen in Fragment A. For example, both forms are seen together in the Melqart Stele, the Amman Citadel Inscription, Panammu I, the Sefir\u00e9 Treaties and the stele of Si\u2019-gabbari from Nerab. In particular, the Sefir\u00e9 Treaties testify to just how divergent forms of letters, especially heh, can be within the one inscription. If there is any question as to different hands being behind the writing of Fragment A and Fragment B, then had the Sefir\u00e9 Treaties been discovered in similarly fragmentary form, scholars would have had grave doubts about any relationship between them. Indeed, the angles of some strokes and the slant of some entire figures in single texts among the Sefir\u00e9 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.
\nThe second difference noted between the fragments was that the \u2018bristles\u2019 in Fragment A tended to become shorter further down the stem. However, this was detected in only three of the six figures of heh in Fragment A (hehs A-2, A-4 and A-5). The other three figures displayed \u2018bristles\u2019 of equal length. Thus, even to begin with, this is not a universal trend. Furthermore, heh B-5, despite damage to the lower half of the letter, displays the top and middle \u2018bristles\u2019 of the figure as being of equal length. Thus, even within Fragment B itself, the tendency to have a shorter middle \u2018bristle\u2019 is not universal. Thus, heh B-5 is for all intents and purposes, identical to one of the more equal \u2018bristled\u2019 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 Sefir\u00e9 Treaties contain a wealth of examples. Another inscription that demonstrates this coexistence of seemingly divergent forms is the Eqron Inscription. There we see figures with a short middle \u2018bristle\u2019 (e.g. line 3), as well as figures with \u2018bristles\u2019 diminishing in size (e.g. line 5). Therefore, the difference in the length of the \u2018bristles\u2019 between Fragment A and Fragment B are certainly not evidence of different hands.
\nThe 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 difference. 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 \u2018bristles\u2019 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.
\nThus, when Fragment A and Fragment B are compared with other Northwest 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 inability 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 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 inscription, or were at least written by the same hand.
\nThis 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\u2019s Horse Ornament, Kilamuwa, the Amman Citadel Inscription, the Nora Stone, the Archaic Phoenician Inscription from Cyprus, Zakkur, Panammu I, the Sefir\u00e9 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.<\/p>\n
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.<\/p>\n
Table 4.25. Occurrences of waw in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 1
\n[\u2026]\u05d5\u05d2\u05d6\u05e8\u05af[\u2026]
\nB-2
\nLine 3
\n\u05d5\u05d9\u05e2\u05dc
\nB-3
\nLine 8
\n[\u2026]\u05d9\u05af\u05d4\u05d5<\/p>\n
The form consists of a vertical stem from which a \u2018hook\u2019 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 \u2018hook\u2019 has the appearance of being round. However, a definite vertex is discernable, meaning the \u2018hook\u2019 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.
\nWaw B-3, however, lacks the vertical stroke of the hook. This peculiarity 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 difficulty that this particular waw (B-3) can be understood as a waw at all. However, there is a good reason why this particular waw is missing half of the \u2018hook\u2019, namely the small pockmark directly adjacent to where we expect the missing stroke to have been. If the stroke had been carved, it would have resulted in an inordinately wide stroke, which was undesirable for aesthetic reasons. The engraver also risked further damage to the written surface through accidental chipping. Instead, this small pockmark was left to act as the vertical stroke of the \u2018hook\u2019. As such, we need not assume Cryer\u2019s reservations over this particular figure.
\nBoth Cryer and Becking cite the letter waw as the most influential palaeographic datum in determining the mutual exclusivity of the Tel Dan fragments. Both point to the angle of the \u2018hook\u2019 as it is displayed in each fragment. In Fragment A, Cryer and Becking claim that the strokes of the \u2018hook\u2019 are perpendicular to each other and branch out perpendicularly from the stem. What is more, Becking claims that the vertical stroke of the \u2018hook\u2019 even leans occasionally inward towards the stem.
\nThis, however, is true for only one waw, namely waw A-1. Most of the other figures do represent this perpendicular \u2018hook\u2019, but there are some notable instances in which this is not the case. The clearest examples are waws A-1 and A-8, which both occur in Line 9. The vertical stroke of the \u2018hook\u2019 in both these figures clearly leans away from the stem towards the left. In this regard, they are identical to the forms of waw in Fragment B. Thus, there is nothing that actually separates the forms within the two fragments. The style of waw used in Fragment B actually matches the scope of the style used in Fragment A.
\nFurthermore, an analysis of other Northwest Semitic inscriptions again reveals these slight variations as normal phenomena that are only to be expected. For example, the Archaic Phoenician Inscription from Cyprus depicts waws with inward facing \u2018hooks\u2019, outward facing \u2018hooks\u2019, as well as perpendicular \u2018hooks\u2019. The same phenomenon is observed on the Kilamuwa Stele, Zakkur, Panammu I, the Sefir\u00e9 Treaties, Panammu II and the Bar Rakib Stele.
\nTherefore, to base the mutual exclusivity of the Tel Dan fragments on the variations of waw is illegitimate. Rather, the waws of both fragments are demonstrably similar. What Cryer and Becking fail to note is that the style of waw used in Fragment B is also used in Fragment A. Also, the strokes of the figures themselves are all proportionate, suggesting a strong association between the fragments rather than a strong disassociation.
\nThe inscriptions mentioned above as typifying the use of slightly varying waws also represent the best matches with the form of Fragment B. Because of the similarity of the styles of waw used in Fragment A, it comes as no surprise that these inscriptions were also the best matches with the waws of Fragment A. This permits us to see a strong Syrian influence over the form, which may be dated between the late ninth century BCE and the late eighth century BCE. The forms are attested in Cyprus, Southeast Anatolia and Syria.<\/p>\n
Zayin. There is one instance of the letter zayin in Fragment B. It occurs in the first line on Fragment B1 and may be seen in Table 4.26 below.<\/p>\n
Table 4.26. Occurrence of zayin in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 1
\n[\u2026]\u05d5\u05d2\u05d6\u05e8\u05af[\u2026]<\/p>\n
The figure is similar to a Latin \u2018H\u2019 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. Similarly, the connector stroke is not exactly vertical but leans to the right. Both parallel strokes are of equal length.
\nSince 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 inscriptions.
\nThe 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 \u2018Z\u2019, becomes the predominant form at the end of the eighth century BCE.
\nOn the A\u1e25iram Sarcophagus, we find a precedent for the form attested here in Fragment B. Two key differences may be noted, though. First, the zayin of A\u1e25iram has relatively short parallel strokes compared to the zayin of Fragment B. Second, the figure in A\u1e25iram characteristically leans to the right. This is typical of an early style, evidenced by other Phoenician inscriptions such as the Shipi\u1e6dba\u02bfal 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.
\nThe 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 inscription is reliable as a measure of other forms. For these reasons, we must refrain from attributing much weight to the match.
\nA 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.
\nThe 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\u2019s Sceptre, in Panammu I and the Sefir\u00e9 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.
\nHowever, with this information, we can make some important observations 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 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\u00b0 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.
\nUnfortunately, 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\u2019s 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.
\nThe 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.<\/p>\n
\u1e24eth. There is only one occurrence of the letter \u1e25eth in Fragment B. It appears on Fragment B1, Line 2. Table 4.27 shows the image of this heth.<\/p>\n
Table 4.27. Occurrence of \u1e25eth in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 2
\n[\u2026]\u05ea\u05af\u05dc\u05d7\u05de\u05d4<\/p>\n
The form of the \u1e25eth here in Fragment B is very similar to that encountered in Fragment A. If anything, the proportion of the \u2018risers\u2019 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 primarily by the \u2018risers\u2019. First, the left \u2018riser\u2019 extends notably above the top \u2018rung\u2019 and only slightly below the bottom \u2018rung\u2019. For the right \u2018riser\u2019 we see the opposite, in which it protrudes only minutely above the top \u2018rung\u2019, but slightly more below the bottom \u2018rung\u2019.
\nCryer\u2019s description of this \u1e25eth is not accurate. In comparing the two forms from both fragments, he states that the form in Fragment A is unusual in displaying the left \u2018riser\u2019 higher than the right \u2018riser\u2019. In making the comparison, he states that such is not the case for the form in Fragment B. Cryer\u2019s statement is somewhat ambiguous in that he does not clarify whether he is referring to the protrusion of the \u2018risers\u2019 beyond the \u2018rungs\u2019, or to the overall latitude at which each \u2018riser\u2019 is placed on the stone. In either case, Cryer\u2019s 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\u2019 respective rotations. The figure in Fragment B leans slightly more leftward than the figure in Fragment A. Yet, both figures display clear rotation. Cryer\u2019s influence on Becking in this regard is unfortunate for both use this inaccurate observation as an argument for two different styles of \u1e25eth 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.
\nThis 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 \u2018rungs\u2019), 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.<\/p>\n
\u1e6ceth. There are no instances of the letter \u1e6deth in Fragment B.<\/p>\n
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.
\nThe forms of yodh here in Fragment B are similar to those found in Fragment A. The two fully preserved figures (yodhs B-1 and B-3) show characteristics of both curved and angular style yodhs. This tendency was also observed on Fragment A. Yodh B-1 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.<\/p>\n
Table 4.28. Occurrences of yodh in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 3
\n\u05d5\u05d9\u05e2\u05dc
\nB-2
\nLine 3
\n\u05de\u05dc\u05db\u05d9\u05af[\u2026]
\nB-3
\nLine 6
\n\u05d0\u05e1\u05e8\u05d9
\nB-4
\nLine 8
\n[\u2026]\u05d9\u05af\u05d4\u05d5<\/p>\n
Since there is such close correspondence between the forms, the same palaeographic considerations for the yodhs 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\u2019s Bowl from Cyprus, the Sefir\u00e9 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.<\/p>\n
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.<\/p>\n
Table 4.29. Occurrences of kaph in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 3
\n\u05de\u05dc\u05db\u05d9\u05af[\u2026]
\nB-2
\nLine 4
\n\u05d4\u05de\u05dc\u05da<\/p>\n
As in Fragment A, the kaph here in Fragment B consists of a long stem from which a \u2018V\u2019 construction branches off to the left, forming what look like three \u2018fingers\u2019 at the top of the figure. There is one slight difference, though. The stem here in Fragment B is notably curved where the form of Fragment A displays only minor curvature. This is probably the point of difference noticed by Becking. The angles at which the \u2018fingers\u2019 protrude are, however, comparable to those in Fragment A.
\nThe closest match with the form of kaph in Fragment B is in Kilamuwa. There we observe that many of the letters have notably curved stems, including the kaph. In all respects, Kilamuwa is an excellent match. Likewise, many of the figures in Mesha, the El-Kerak Fragment and the Amman Citadel Inscription are very close to the form here in Fragment B. From the same period, a votive inscription on a stone vessel from Kuntilet \u2018Ajrud also bears resemblance with the form in Fragment B; only the \u2018fingers\u2019 are not quite as spread apart. The same style is observed on the Ivory Plaque found at Nimrud, the Siloam Tunnel Inscription, and the Eqron Inscription. Some examples from the Azitawadda Inscription also qualify for consideration, thus yielding us a chronological bracket between the mid-ninth century BCE and the early seventh century BCE. The forms are attested in Southeast Anatolia, Syria, Transjordan, Palestine and the Negev.
\nIf we discount the curvature of the stem, many more inscriptions become available for comparison. However, these inscriptions all fall within the same chronological bracket as the previously mentioned matches\u2014namely, the mid-ninth centiuy BCE through to the early seventh century BCE. Yet, to incorporate non-curving matches is to deny the clear style of kaph in Fragment B.
\nWe must now ask whether the two forms, represented by Fragment A and Fragment B respectively, are at all reconcilable. Since the \u2018fingers\u2019 of both figures match up, the issue inevitably comes down to the significance of the curved stem. In order to ascertain the importance of this, we must use the evidence from other inscriptions.
\nFor this task, the Mesha Stele is most informative. On the stele we find a clear shift in the curvature of the stem for the letter kaph. In the first 16 lines of text, only three of the legible kaphs display a curved stem comparable to the figures in Fragment B, while 23 figures have stems either comparable to Fragment A or straighter. At line 17, a transition occurs so that in lines 17 through to 34, we see twenty curved-stem kaphs comparable to Fragment B and only nine stems of kaphs comparable to Fragment A. In the last ten lines of text (line 25\u201334), almost all the kaphs have very curved stems.
\nIn practical terms, these data from the Mesha Stele demonstrate that in the upper portions of a stele, kaphs are more likely to have straighter stems than in the bottom portions. The transition between the two forms occurs precisely at a point where a scribe\u2019s arm would not be required to stretch inordinately forward. Thus, in the bottom portions when a scribe does not need to stretch so far forward to write, more comfort in writing is achieved and therefore more stylistic strokes, such as curved stems, are achievable. This is also generally the case with the Sefir\u00e9 Treaties, though the problem there is that the kaphs are clearly meant to possess straighter rather than curved stems.
\nThis means that the seemingly different forms of kaph in the Tel Dan fragments are certainly reconcilable, especially if Fragment B is to be placed below Fragment A, rather than to its left as it was arranged by Biran, Naveh and their three colleagues. Therefore, the difference in styles may be interpreted in either way, as evidence for or against the integrity of the fragments as one inscription.<\/p>\n
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.<\/p>\n
Table 4.30. Occurrences of lamed in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 2
\n[\u2026]\u05ea\u05af\u05dc\u05d7\u05de\u05d4
\nB-2
\nLine 3
\n\u05d5\u05d9\u05e2\u05dc
\nB-3
\nLine 3
\n\u05de\u05dc\u05db\u05d9\u05af[\u2026]
\nB-4
\nLine 4
\n\u05d4\u05de\u05dc\u05da<\/p>\n
All of the figures are in the shape of a \u2018fish hook\u2019. The \u2018shanks\u2019 show little or no curvature, while the actual hook is a neat curve. There is evidence of a slight \u2018bump\u2019 in the \u2018hook\u2019 of lamed B-2. The \u2018barbs\u2019 on all figures slant noticeably upwards, pointing essentially in the same direction as the top of the \u2018shank\u2019.
\nBecking notices a difference between the form here in Fragment B and the form represented in Fragment A. He does not, however, specify the nature of the difference. We can only surmise that it is the uniform curvature of the \u2018hook\u2019 that Becking has in mind. The figures attested on Fragment A do not display such uniform roundness at the \u2018hook\u2019, but are prone to hyperbolic roundness which makes the \u2018hook\u2019 look like a distinct bend or \u2018bump\u2019 rather than a smooth curve.
\nHowever, lamed B-2 also shows evidence of this \u2018bump\u2019 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 \u2018hook\u2019 so that it is similar to the forms on Fragment B. Thus, there is adequate overlap of the forms on both fragments, and so the figures do not lend support to Becking\u2019s claim of exclusive differences.
\nSince 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 A\u1e25iram, Ye\u1e25imilk and Shipi\u1e6dba\u02bfal. The Mesha Stele forms are comparable in all aspects except the \u2018barbs\u2019, 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 Inscription, the Archaic Phoenician Inscription from Cyprus, Zakkur, Panammu I, the Sefir\u00e9 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 \u2018yahu ben-Sanballa\u1e6d, Governor of Samaria.
\nThus, the data on the form of lamed gives us a huge chronological horizon stretching six centuries\u2014from 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.<\/p>\n
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.<\/p>\n
Table 4.31. Occurrences of mem in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 2
\n[\u2026]\u05ea\u05af\u05dc\u05d7\u05de\u05d4
\nB-2
\nLine 3
\n\u05de\u05dc\u05db\u05d9\u05af[\u2026]
\nB-3
\nLine 4
\n\u05d4\u05de\u05dc\u05da
\nB-4
\nLine 5
\n\u05de\u05df\u05af
\nB-5
\nLine 7
\n[\u2026]\u05e8\u05dd<\/p>\n
There is a slight difference in the forms here in Fragment B. The first three figures (mems B-1, B-2, B-3) have a continuous \u2018jagged head\u2019 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 \u2018jagged head\u2019 do not meet end to end. Rather, it appears that the left two strokes of the \u2018jagged head\u2019 have slipped down slightly.
\nThis 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 characteristic markings of the form were the angle of the leftmost stroke of the \u2018jagged head\u2019 and the angle of the \u2018jagged head\u2019 as a whole. When we compare these points with those of the forms here in Fragment B, we observe 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 \u2018jagged head\u2019 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.
\nThe 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 be discussed further in the synthesis of all the epigraphical and palaeographical information.
\nSince the forms of mem in both Fragment A and Fragment B are virtually identical, we have the exact same list of matches from other North-west Semitic inscriptions for both fragments. Thus, the best matches for the mem of Fragment B are the Mesha Stele, the El-Kerak Fragment, Kilamuwa, the Amman Citadel Inscription, Zakkur, Panammu I, the Sefir\u00e9 Treaties, the inscribed bricks from Hamath, Panammu II and Bar Rakib.
\nWith these data, we may assign the mem of Fragment B to the time between the mid-ninth century BCE and the late eighth century BCE. The matches are attested in Southeast Anatolia, Syria and Transjordan.<\/p>\n
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.<\/p>\n
Table 4.32. Occurrences of nun in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 5
\n\u05de\u05df\u05af
\nB-2
\nLine 6
\n[\u2026]\u05e2\u05af\u05df\u05af<\/p>\n
The form of nun in Fragment B is difficult to assess because of the fragmentary nature of both attestations. We may piece the portions of each example to gain a coherent form, but this would be only hypothetical. This hypothetical form certainly matches the two forms encountered in Fragment A. Whereas the stem is completely destroyed on nun B-2, the very bottom portion still remains nun B-1, preserved below the join of Fragments B1 and B2. This small portion allows the observation that at least the stem appears to be curved in a way similar to the figures attested in Fragment A.
\nA definitive comparison with the forms of Fragment A is, however, elusive because of the fragmentary nature of nun B-1. This makes it difficult to reach a verdict on the relationship with the forms of Fragment A. For this same reason, a palaeographical comparison with other Northwest 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. Comprehensive comparison, however, cannot be made.<\/p>\n
Samekh. There is a single occurrence of the letter samekh in Fragment B, the image of which is seen below in Table 4.33.<\/p>\n
Table 4.33. Occurrences of samekh in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 6
\n\u05d0\u05af\u05e1\u05e8\u05d9<\/p>\n
The form consists of a slightly curved, slightly leftward leaning stem, which is intersected at the top by three crossbars. The form is the same as that encountered in Fragment A. The only difference that might be perceived between the two forms is the length of the stem. However, this is not a real difference because the figures in the top half of Fragment A are discernably bigger than those in the second half of Fragment A and the figures in Fragment B. Indeed, both figures of samekh are proportionally identical.
\nWith these observations, it may plausibly be suggested that the same hand is at work in both figures. This means the matches that were cited for the form in Fragment A will also apply here for Fragment B. Thus, relevant matches of a slightly leftward leaning samekh are made with forms from Kilamuwa\u2019s Sceptre, the Amman Citadel Inscription, the Hamath Bricks and the Azitawadda Inscription. This yields us forms of Syrian style from the late ninth century BCE through to the late eighth century BCE. The forms are attested in Southeast Anatolia, Syria and Transjordan.<\/p>\n
\u2018Ayin. There are three instances of the letter \u2018ayin in Fragment B. Their images are seen below in Table 4.34.
\nAs 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 here\u2014namely, 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 \u2018bump\u2019 at the bottom left of \u2018ayin B-1. Although this is not visible on \u2018ayin B-2, the curvature at the bottom of the letter hints at such a \u2018bump\u2019.<\/p>\n
Table 4.34. Occurrences of \u2018ayin in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 3
\n\u05d5\u05d9\u05e2\u05dc
\nB-2
\nLine 5
\n\u05e9\u05af\u05d1\u05e2\u05af[\u2026]
\nB-3
\nLine 6
\n[\u2026]\u05e2\u05af\u05df\u05af<\/p>\n
In his description of the difference in the \u2018ayins between the fragments, Cryer states that the forms in Fragment A are ovoid shapes, whereas the single fully preserved form in Fragment B is round. This, however, is not a realistic difference, nor even a discerning observation. The characteristic \u2018bumps\u2019 along the curvature of the letters are observed in both fragments. This speaks for both the same hand and the same engraver lying behind both Fragment A and Fragment B. Also, Cryer is much too stringent in his allowance for a margin of variation between separate figures. When the \u2018ayins of both fragments are lined up, they are seen to be virtually identical in both shape and size while also demonstrating that the human hand is not mechanically precise in all minutiae. More importantly, the characteristic features of each form are the same. Thus, Cryer\u2019s objection to a match between the \u2018ayins of the two sets of fragments cannot be upheld.
\nThe consistency in both forms is further demonstrated by the fact that the matches for the form in Fragment B are exactly the same as those for the form in Fragment A. We notice that the smaller form of \u2018ayin is also employed in Fragment B. Good matches are found with the Mesha Stele (mid-ninth century BCE) at the earliest, and the Eshmun\u2018azar Sarcophagus (early fifth century BCE) at the latest. Between these, we find correspondence with forms in the Amman Citadel Inscription, the Archaic Phoenician Inscription from Cyprus, the inscriptions of Panammu I, Panammu II, the Nerab Stelae, the Siloam Tunnel, the Silwan Tomb, the Eqron Inscription and the Sarcophagus of Tabnit.
\nThus, for the letter \u2018ayin, we find the form in use across the entire Levant, as well as Cyprus, between the mid-ninth century BCE and the early fifth century BCE.<\/p>\n
Peh. There is a single occurrence of the letter peh in Fragment B. Its image is shown below in Table 4.35.<\/p>\n
Table 4.35. Occurrences of pe in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 5
\n[\u2026]\u05d0\u05e4\u05e7<\/p>\n
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-1, suggesting a connection between the two forms.
\nSince there is such a close correlation between the forms of both fragments, 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 mainland, matches are found in the Zakkur Stele, the Sefir\u00e9 Treaties, Azitawad-da, Panammu II, Bar Rakib and the sarcophagi of Tabnit and Eshmun\u2018azar.
\nThese data provide us a chronological bracket identical to the form in Fragment A\u2014namely, between the late ninth century BCE and the early fifth century BCE. The form is attested in Sardinia, Cyprus, Southeast Anatolia, Syria and Phoenicia.<\/p>\n
\u1e62adhe. There are no occurrences of the letter \u1e62adhe in Fragment B.<\/p>\n
Qoph. There is one instance of the letter qoph in Fragment B. Its image is shown below in Table 4.36.
\nThe 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.<\/p>\n
Table 4.36. Occurrences of qoph in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 5
\n[\u2026]\u05d0\u05e4\u05e7<\/p>\n
Once again, the fine correspondence between the forms means our matches for the form in Fragment A are the same for Fragment B. Therefore, 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\u2019s Horse Ornament, both Kilamuwa\u2019s Stele and Sceptre, the Archaic Phoenician Inscription from Cyprus, Zakkur, Panammu I, the Sefir\u00e9 Treaties, Azitawadda, Panammu II and the Nerab Stelae.
\nTherefore, 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.<\/p>\n
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.<\/p>\n
Table 4.37. Occurrences of resh in Fragment B<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 1
\n[\u2026]\u05d5\u05d2\u05d6\u05e8\u05af[\u2026]
\nB-2
\nLine 6
\n\u05d0\u05af\u05e1\u05e8\u05d9
\nB-3
\nLine 7
\n[\u2026]\u05e8\u05dd
\nB-4
\nLine 7
\n\u05d1\u05e8
\nB-5
\nLine 8
\n\u05d1\u05e8[\u2026]<\/p>\n
As in Fragment A, the form consists of a long stem and a neat triangular \u2018head\u2019, with the whole figure being tilted to the left. There is, however, a noticeable difference with Fragment A in the shape of the \u2018head\u2019. This was marked by both Cryer and Becking. Whereas the \u2018head\u2019 on the figures in Fragment A are quite large, they are quite small and even squat in Fragment B. To put the difference another way, the interior angles of the triangular \u2018head\u2019 differ between the fragments. This difference in angles also makes the stem of the figures in Fragment B look longer than those in Fragment A.
\nWhen the overall sizes of the figures from both Fragments are compared, there is no real difference. In fact, the stems in both fragments have a virtually identical range in length. Similarly, the \u2018heads\u2019 from both fragments have a virtually identical range in width. That is, the horizontal stroke of the \u2018head\u2019 is roughly the same length in the figures from both fragments.
\nThus, although differences between the forms in the two fragments are certainly observable, they are by no means major. While there is the suggestion that the differences betray the fragments as two different inscriptions, the evidence certainly does not demand it. On the contrary, the forms are comparable enough to even suggest the same hand behind both of them. The Archaic Phoenician Inscription from Cyprus uses reshs similar to Fragment A in the first few lines, whereas the form used in the last extant line is identical to the form of Fragment B. Also, both the stele of Panammu I and the Sefir\u00e9 Treaties use both forms, indicating that they are a variation of the same style. As such, the comments of Cryer and Becking are far from persuasive.
\nIn finding matches for the form of resh in Fragment B, we must look for comparably slanted figures rather than upright ones. The earliest inscription with matching forms is the Melqart Stele. Some forms at the bottom of the Mesha Stele also match. We may also cite examples from both Kilamuwa\u2019s Stele and Sceptre, the Amman Citadel Inscription, the previously mentioned example from the Archaic Phoenician Inscription from Cyprus, Panammu I, the Sefir\u00e9 Treaties, the Hamath Bricks, Panammu II and, to a lesser degree, Bar Rakib. The sarcophagi of Tabnit and Eshmun\u2018azar also qualify as matches.
\nThus, 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 fragments 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.<\/p>\n
\u0160in. There is a single instance of the letter \u0161in in Fragment B. It occurs right on the join between Fragment B1 and Fragment B2. Its image is contained in Table 4.38 below.<\/p>\n
Table 4.38. Occurrence of \u0161in in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 5
\n\u05e9\u05af\u05d1\u05e2\u05af[\u2026]<\/p>\n
Despite the damage to the figure from the fracturing of the stone, the form is essentially the same as that encountered in Fragment A. The forms match up in four ways. First, the leftmost stroke leans to the left. This was seen as a feature allowing us to distinguish a \u0161in from a mem in a damaged context. Second, the right \u2018V\u2019 is slightly larger than the left \u2018V\u2019. Third, the right \u2018V\u2019 sits slightly lower than the left \u2018V\u2019. Fourth, the left \u2018V\u2019 is rotated in more of a clockwise direction than the right \u2018V\u2019. These points of correspondence make a strong case for the same hand lying behind both Fragment A and Fragment B.
\nDue to this close correspondence, the matches found for the form in Fragment A also apply here for Fragment B. As such, we find matches for the Fragment B \u0161in in the Tell Fakhariyah Inscription, the Amman Citadel Inscription, Nora, Zakkur, Panammu I, the Phoenician Governor\u2019s Bowl from Cyprus, the Sefir\u00e9 Treaties, Azitawadda, Panammu II, Bar Rakib, the Nerab Stelae and the Eqron Inscription.
\nThese data provide us with a chronological horizon between the mid-ninth century BCE and the early seventh century BCE. The form is attested in Sardinia, Cyprus, Southeast Anatolia, Syria, Palestine, Transjordan and Mesopotamia.<\/p>\n
Taw. There is only one instance of taw in Fragment B. It is, however, mostly damaged and has been restored on epigraphical and philological grounds. The remnants of this taw are seen below in Table 4.39.<\/p>\n
Table 4.39. Occurrence of taw in Fragment B.<\/p>\n
No.
\nImage
\nLocation
\nOccurring Word
\nB-1
\nLine 2
\n[\u2026]\u05ea\u05af\u05dc\u05d7\u05de\u05d4<\/p>\n
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-1, 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.
\nSince 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, however, note that the form is a long-stemmed taw, rather than the shorter equilateral taw. This shows a clear Syrian influence over the form of the letter.<\/p>\n
Dating the Script of Fragment B
\nAs was the case with the script of Fragment A, we see a decidedly Syrian influence over the script of Fragment B. The tendency to lean letters to the left: is probably the most telling characteristic. The time-frame common to each individual letterform attested in the fragment is that between the late ninth century BCE and the late eighth century BCE. In script development, this corresponds with the inscriptions between the Kilamuwa Stele and the Bar Rakib Stele or Azitawadda Inscription. These dates correspond with that gleaned from the archaeological context of the fragments.
\nOf all the letters, the heth provides us with perhaps the best evidence to work towards narrowing this bracket. We observed how the \u1e25eth of Fragment B displayed all the characteristic features of the \u1e25eth of Fragment A. 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 \u1e25eth is one of the most complex letters, it is a fairly accurate indicator of variant styles.
\nThe forms of \u2019aleph and \u2018ayin 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 A\u2014namely, c. 800 BCE (\u00b1 20 years).<\/p>\n
Synthesis of the Script of Fragment A and the Script of Fragment B<\/p>\n
The date arrived at for the script of Fragment B (c. 800 BCE, +\/\u2212 20 years) is confirmed by the fact that the script of Fragment B is very close, if not identical, to the script of Fragment A. If we perform a statistical analysis of the relationship between the fragments for each individual letter, we see this fact substantiated. First, four letters (gimel, zayin, teth and \u1e63adhe) are either attested in only one fragment or in neither. Therefore, we may dismiss them for purposes of considering a relationship between the fragments. Of the remaining 18 letters, ten (56 per cent) show a very close correspondence and six (33 per cent) appear to be closely related variants. The remaining two (11 per cent) letters (nun and taw) are special cases in that they are incomplete figures in Fragment B and difficult to compare with their counterparts in Fragment A. Most importantly, no letters are irreconcilably different.
\nThis last particular point may legitimately be claimed for many contemporary but mutually exclusive inscriptions. As such, the fact that no letters are irreconcilably different is no guarantee that the fragments are to be understood as belonging to the same inscription. However, it does much to undermine the objections of Cryer, Becking and others who claim that the difference between the fragments demand the interpretation that the fragments are mutually exclusive. This is certainly not the case.
\nFurthermore, two particular letters are invaluable in actually clinching a relationship between Fragment A and Fragment B. First, the letter waw is noted by both Cryer and Becking as probably the most obvious and telling figure for the mutual exclusivity of the fragments. However, close observation of the forms in both fragments reveals that the forms do in fact overlap quite significantly. As such, the letter waw provides a bridge rather than a chasm between Fragment A and Fragment B.
\nSecond, 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 \u2018jagged head\u2019 had slipped slightly downwards. This internal inconsistency alerts us to the tendency of the scribe of Fragment B to make slight alterations 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 \u2018bristles\u2019 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 established. 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 palaeographical analysis also confirms the veracity of the date for the fragments yielded in our examination of the archaeological context. As such, we may legitimately claim that the Tel Dan Inscription was written at some time close to 800 BCE.<\/p>\n
Summary of Script Analysis<\/p>\n
Tables 4.40 and 4.41 summarize the palaeographical information gleaned on each set of letters from both Fragment A and Fragment B.<\/p>\n
Table 4.40. Summary of Palaeographical Information for the Script of Fragment A.<\/p>\n
FRAGMENT A
\nLetter
\nClosest Matching Inscriptions
\nRegional Bracket
\nChronological Bracket (centuries BCE)
\n\u05d0
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nHamath Bricks
\nNerab Stelae
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia Syria
\nTransjordan
\nLate ninth\u2013early fifth
\n\u05d1
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nEqron
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nTema
\nNorthwest Arabia
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate ninth\u2013mid-fifth
\n\u05d2
\n\u2014
\n\u2014
\n\u2014
\n\u05d3
\nKilamuwa
\nAmman Citadel
\n(Samaria Ostraca)
\nHamath Bricks
\nBar Rakib
\nSiloam Tunnel
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate-ninth\u2013late eighth
\n\u05d4
\nMelqart Stele
\nMesha Stele
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nHazael\u2019s Horse Ornament
\nZakkur
\nBar Rakib
\nNerab Stelae
\nEqron
\nCyprus
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05d5
\nKilamuwa
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nBar Rakib
\nCyprus
\nSoutheast Anatolia
\nLate ninth\u2013late-eighth
\n\u05d6
\n\u2014
\n\u2014
\n\u2014
\n\u05d7
\nTell Fakhariyah
\nHazael Ivory
\nKilamuwa
\nAmman Citadel
\nZakkur
\nSiloam Tunnel
\nPalestine
\nSamal
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05d8
\n\u2014
\n\u2014
\n\u2014
\n\u05d9
\nMelqart Stele
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nAzitawadda
\nBar Rakib
\nNimrud Ivory Plaque
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05db
\nMesha Stele
\nKuntilet \u2018Ajrud
\nAmman Citadel
\nNora Inscription
\nPhoenician Governor\u2019s Bowl
\nPanammu I
\nHamath Bricks
\nBar Rakib
\nSiloam Tunnel
\nEqron
\nCyprus
\nNegev
\nPalestine
\nSE Anatolia
\nSardinia
\nSyria
\nTransjordan
\nLate ninth\u2013early seventh
\n\u05dc
\nA\u1e25iram Sarcophagus
\nYe\u1e25imilk shipi\u1e6dba\u02bfal
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nKuntilet \u2018Ajrud Wares
\nZakkur
\n(Samaria Ostraca)
\n(Samaria Seals)
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\n(Khirbet el-Qom)
\n(Ophel Ostracon)
\nAzitawadda
\nPanammu II
\nBar Rakib
\n(Tel Qasil\u00e9 Ostracon II)
\n(Beersheba Juglet)
\nNerab Stelae
\n(Lachish III Seals)
\n(Ramat Ra\u1e25el Seal)
\n(Tell en-Na\u1e63beh Seal)
\n(Wadi Muraba\u2018at Papyrus)
\n(Shechem III Seal)
\nTema
\n(Yesha\u2018yahu Bulla)
\nCyprus
\nNegev
\nNorthwest Arabia
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-tenth\u2013mid-fourth
\n\u05de
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nBar Rakib
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05e0
\nMelqart Stele
\nMesha Stele
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nSa\u1e43al
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05e1
\nKilamuwa Scepter
\nAmman Citadel
\nHamath Bricks
\nAzitawadda
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate ninth\u2013late eighth
\n\u05e2
\nMesha Stele
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nPanammu II
\nNerab Stelae
\nSiloam Tunnel
\nSilwan Tomb
\nEqron
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early fifth
\n\u05e4
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nLate ninth\u2013early fifth
\n\u05e6
\nComparison not practicable
\n\u2014
\n\u2014
\n\u05e7
\nMelqart Stele
\nHazael\u2019s Horse Ornament
\nKilamuwa
\nKilamuwa\u2019s Sceptre
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nMid-ninth\u2013early seventh
\n\u05e8
\nMelqart Stele
\nMesha Stele
\nKilamuwa
\nKilamuwa Sceptre
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nTema
\nCyprus
\nNorthwest Arabia
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013mid-fifth
\n\u05e9
\nTell Fakhariyah
\nAmman Citadel
\nNora
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nNerab Stelae
\nEqron
\nCyprus
\nMesopotamia
\nPalestine
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05ea
\nMelqart Stele
\nHazael Ivory
\nHazael\u2019s Horse Ornament
\nArchaic Phoenician (Cyprus)
\nKilamuwa
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nAzitawadda
\nPanammu II
\nBar Rakib
\nNerab Stelae
\nCyprus
\nSoutheast Anatolia
\nSyria
\nMid-ninth\u2013late eighth
\nDate for script of Fragment A: c. 800 BCE (\u00b1 20 years)<\/p>\n
Table 4.41. Summary of Palaeographical Information for the Script of Fragment B.<\/p>\n
FRAGMENT B
\nLetter
\nClosest Matching Inscriptions
\nRegional Bracket
\nChronological Bracket (centuries BCE)
\n\u05d0
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nHamath Bricks
\nNerab Stelae
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nLate ninth\u2013early fifth
\n\u05d1
\nMelqart Stele
\nKilamuwa
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nBar Rakib
\nNerab Stelae
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\n(Tema)
\nCyprus
\n(Northwest Arabia)
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nLate ninth\u2013early fifth
\n\u05d2
\nYe\u1e25imilk
\nEliaba\u2018al
\nKilamuwa
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nPanammu I
\nPanammu II
\nSefir\u00e9 Treaties
\nBar Rakib
\nNerab Stelae
\nTabnit Sarcophagus
\nTema
\nCyprus
\nNorthwest Arabia
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nLate tenth\u2013mid-fifth
\n\u05d3
\nKilamuwa
\nAmman Citadel
\n(Samaria Ostraca)
\nHamath Bricks
\nBar Rakib
\nSiloam Tunnel
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate-ninth\u2013late eighth
\n\u05d4
\nMelqart Stele
\nHazael\u2019s Horse Ornament
\nKilamuwa
\nAmman Citadel
\nNora Inscription
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nBar Rakib
\nSilwan Tomb
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05d5
\nKilamuwa
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nPanammu II
\nBar Rakib
\nCyprus
\nSoutheast Anatolia
\nSyria
\nLate ninth\u2013late eighth
\n\u05d6
\nMelqart Stele
\nKilamuwa
\nKilamuwa\u2019s Sceptre
\nPanammu I
\nAzitawadda
\nSefir\u00e9 Treaties
\nSiloam Tunnel
\nSilwan Tomb
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05d7
\nTell Fakhariyah
\nHazael Ivory
\nKilamuwa
\nAmman Citadel
\nZakkur
\nSiloam Tunnel
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05d8
\n\u2014
\n\u2014
\n\u2014
\n\u05d9
\nMelqart Stele
\nKilamuwa
\nAmman Citadel Inscription
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nBar Rakib
\nAzitawadda
\nNimrud Ivory Plaque
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05db
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel Inscription
\nKuntilet \u2018Ajrud
\nNimrud Ivory Plaque
\nAzitawadda
\nSiloam Tunnel
\nEqron
\nNegev
\nPalestine
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013early seventh
\n\u05dc
\nA\u1e25iram Sarcophagus
\nYe\u1e25imilk
\nShipi\u1e6dba\u02bfal
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nArchaic Cyprus
\nNora Inscription
\nKuntilet \u2018Ajrud Wares
\nZakkur
\n(Samaria Ostraca)
\n(Samaria Seals)
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\n(Khirbet el-Qom)
\n(Ophel Ostracon)
\nAzitawadda
\nPanammu II
\nBar Rakib
\n(Tel Qasil\u00e9 Ostracon II)
\n(Beersheba Juglet)
\nNerab Stelae
\n(Lachish III Seals)
\n(Ramat Ra\u1e25el Seal)
\n(Tell en-Na\u1e63beh Seal)
\n(Wadi Muraba\u2018at Papyrus)
\n(Shechem III Seal)
\nTema
\nYesha\u2018yahu Bulla
\nCyprus
\nNegev
\nNorthwest Arabia
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-tenth\u2013mid-fourth
\n\u05de
\nMesha Stele
\nEl-Kerak Fragment
\nKilamuwa
\nAmman Citadel
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nBar Rakib
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nMid-ninth\u2013late eighth
\n\u05e0
\nComparison not practicable
\n\u2014
\n\u2014
\n\u05e1
\nKilamuwa Sceptre
\nAmman Citadel
\nHamath Bricks
\nAzitawadda
\nSoutheast Anatolia
\nSyria
\nTransjordan
\nLate ninth\u2013late eighth
\n\u05e2
\nMesha Stele
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nPanammu II
\nNerab Stelae
\nSiloam Tunnel
\nSilwan Tomb
\nEqron
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPalestine
\nPhoenicia
\nSoutheast Anatolia
\nSyria
\nTransjorden
\nMid-ninth\u2013early fifth
\n\u05e4
\nArchaic Phoenician (Cyprus)
\nNora Inscription
\nZakkur
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSardinia
\nSyria
\nLate ninth\u2013early fifth
\n\u05e6
\n\u2014
\n\u2014
\n\u2014
\n\u05e7
\nMelqart Stele
\nHazael\u2019s Horse Ornament
\nKilamuwa
\nKilamuwa Sceptre
\nArchaic Phoenician (Cyprus)
\nZakkur
\nPanammu I
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nNerab Stelae
\nEqron
\nCyprus
\nPalestine
\nSoutheast Anatolia
\nSyria
\nMid-ninth\u2013early seventh
\n\u05e8
\nMelqart Stele
\nMesha Stele
\nKilamuwa
\nKilamuwa Sceptre
\nAmman Citadel
\nArchaic Phoenician (Cyprus)
\nPanammu I
\nSefir\u00e9 Treaties
\nHamath Bricks
\nPanammu II
\nBar Rakib
\nTabnit Sarcophagus
\nEshmun\u2018azar Sarcophagus
\nCyprus
\nPhoenicia
\nSoutheast Anatolia
\nSyria
\nTransjorden
\nMid-ninth\u2013mid-fifth
\n\u05e9
\nTell Fakhariyah
\nAmman Citadel
\nNora Inscription
\nZakkur
\nPanammu I
\nPhoenician Governor\u2019s Bowl
\nSefir\u00e9 Treaties
\nAzitawadda
\nPanammu II
\nBar Rakib
\nNerab Stelae
\nEqron
\nCyprus
\nMesopotamia
\nPalestine
\nSoutheast Anatolia
\nSardinia
\nSyria
\nTransjordan
\nMid-ninth-early seventh
\n\u05ea
\nComparison not practicable
\nSyrian Influence
\n\u2014
\nDate for script of Fragment B: c. 800 BCE (\u00b1 20 years)<\/p>\n
Chapter 5<\/p>\n
ARRANGEMENT OF THE FRAGMENTS<\/p>\n
Evaluation of the Original Arrangement by Biran and Naveh<\/p>\n
Introductory Remarks
\nIn this section, I will attempt a re-evaluation of the arrangement of the fragments. Certain observations already made in the epigraphical analysis of the fragments suggest there is doubt as to the arrangement endorsed by Biran and Naveh. This now needs to be assessed.
\nWhen Biran and Naveh published Fragment B, they offered a text based on a particular arrangement of the fragments. This saw Fragment B placed to the left of Fragment A, with the first extant line of text in Fragment A lined up with the first extant line of text in Fragment B (Fig. 5.1). Biran and Naveh mention that three different arrangements were considered. The first was to place Fragment B above Fragment A\u2014a configuration rejected by Biran and Naveh on the grounds of literary genre. The publishers claimed that references to an author\u2019s father in an inscription traditionally occurred in the first few lines of an inscription. Since Fragment A made mention of the author\u2019s father, Biran and Naveh were reluctant to place the eight lines of text represented by Fragment B prior to it.
\nThe second arrangement considered by Biran and Naveh was to place Fragment B after Fragment A. No comment was made as to the suitability of this configuration. Rather, it was passed up in favour of the third arrangement, which placed Fragment B to the left of Fragment A. The reasons for this were textual. According to Biran and Naveh, this \u2018placement provided, albeit with some difficulties, a meaningful and continuous text\u2019.<\/p>\n
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).<\/p>\n
This arrangement was further endorsed by three restorers who worked independently to ascertain whether Fragment A and Fragment B could be joined together in any way. These three restorers were Nili Cohen of the Hebrew Union College, and A. Weiner and Ruth Yekutieli, both with the Israel Museum. According to Biran and Naveh, each of these three restorers produced the same arrangement, demonstrating that the two fragments could be joined under the surface at Line 5. In achieving this join, the bottom left edge of Fragment A and the top right edge of Fragment B were seen as part of the same line of fracture. According to Ada Yardeni, who was also instrumental in producing this arrangement, this join under the surface was more than satisfactory.
\nThus, the fragments of the Tel Dan Inscription were presented to the scholarly community and the world with Fragment B placed to the immediate left of Fragment A. It is in this same configuration that people may now view the inscription, housed in a glass case, in the Archaeology Section of the Israel Museum in Jerusalem.
\nIn their comments on the arrangement of the fragments, Biran and Naveh stated that \u2018Fragments A and B cannot be joined in an obvious, unequivocal way\u2019. Thus, by their own admission, there is a modicum of doubt as to the present configuration. Biran reiterated these comments at a later stage. The convenience of the join must also be doubted, for Biran and Naveh also stated that the text that results is not without its problems.
\nThus, there is reason to investigate the integrity of the arrangement proposed by Biran, Naveh and their colleagues, Yardeni, Cohen, Weiner and Yekutieli. Doubts have also been raised in the foregoing epigraphical analysis of the fragments. Therefore, we must attempt an evaluation to determine the strength of the original arrangement. This involves pursuing those points of doubt that arise from the arrangement. The fact that the current configuration is preferred over any other, because it provides a meaningful text, means that investigation of the configuration is of paramount importance. If the arrangement is seen to be faulty in any way, then this will have direct implications for the reading of the text.<\/p>\n
The Physical Join
\nThe join between Fragment A and Fragment B beneath the surface of Lines 4 and 5 does not present an easy interlocking fit. The right extremity of Fragment B may be placed up against part of the exposed surface of Fragment A at the latitudes of Lines 4 and 5 (see Fig. 5.1). The length of the join is approximately 3 cm. However, the fragments do not slip together naturally and there are gaps in this \u2018join\u2019. The nature of the join is similar to two clenched fists placed end to end; they sit against each other but do not really fit into each other. As Yamada phrases it, the \u2018physical [sic join between Fragments A and B1+B2 is based on a quite infirm point of contact\u2019.
\nSome of the exposed surfaces in the upper reaches of both fragments come close to joining but do not manage to actually touch. One of the strengths of this original arrangement, however, is the fact that there appears to be one continuous line of fracture running from the bottom left of Fragment A up towards the top of Fragment B. Similarly, there is an almost perpendicular line of breakage running from the top left of Fragment A, down to Line 5, and then down along the bottom right of Fragment B1. However, since none of the written surfaces join up, there must be some doubt as to whether the line of breakage of one fragment is actually the same as the line on the other fragment. Thus, physically, the join that Biran, Naveh, and their colleagues propose is possible, but it cannot be confirmed.<\/p>\n
Epigraphical Considerations
\nDoubts about the integrity of the original arrangement were first raised by Cryer and Thompson of the University of Copenhagen in two articles published in the Scandinavian Journal of the Old Testament. This occurred immediately following the publication of Fragment B in the Israel Exploration Journal by Biran and Naveh. Cryer\u2019s 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.
\nThompson\u2019s objections, which led him to the same conclusion, were based primarily on archaeological and epigraphic grounds. Thompson questioned whether Fragment A and Fragment B had a common archaeological context seeing as they were discovered in different usages at different locations 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 accommodate certain letters proposed by Biran and Naveh between Fragment A and Fragment B.
\nI will deal with Thompson\u2019s 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 Fragment B if it was originally located to the left of Fragment A. We would expect to see this trend continue in Fragment B. Yet, the trend of the lines in Fragment B shows a completely different tendency. The trend of the lines there indicates that the scribe was nearing the bottom of the stone and so did not need to stretch forward to write at all. Thus, the misalignment of the lines is confirmed by the vastly different trends detectable in the lines of each fragment.<\/p>\n
Figure 5.2. The incongruence of line trends between Fragment A (right) and Fragment B (left) (Tel Dan Excavations, Hebrew Union College, Jerusalem; photograph: Z. Radovan [detail]).<\/p>\n
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. 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 fragments. He rotated Fragment B1 2\u00b0 in an anti-clockwise direction and Fragment B2 1.5\u00b0 in the same direction. Thus, Fragments B1 and B2 were offset by 0.5\u00b0. According to Schniedewind, this \u2018slight rotation of the fragments makes the lines match better and renders a more convincing join\u2019.
\nHowever, numerous objections must be raised to Schniedewind\u2019s proposal 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 rotation of the fragments is physically possible.
\nNevertheless, Schniedewind\u2019s claim that the rotation produces a better alignment and more convincing join between the fragments cannot be substantiated. 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\u2019s proposal does not bring us any closer to substantiating the join proposed by Biran and Naveh et al.
\nIt is clear from the trends of the lines that different physical circumstances 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.
\nMore 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 \u2018head\u2019 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. I 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 consistent with a scribe who reached the bottom portion of the stone and needed to cram his letters slightly.<\/p>\n
Palaeographical Considerations
\nThe slight difference in the forms of some letters led Cryer to claim the mutual exclusivity of Fragment A and Fragment B. Others, like Becking, soon followed with similar hypotheses. In the earlier palaeographical analysis of the fragments, we saw how these differences were not sufficient to warrant the interpretation that we were dealing with two separate inscriptions. 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).
\nHowever, 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\u2019s arm must have been in a very comfortable position during writing of the lines in Fragment B.
\nIf 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\u2019s 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 movement evident in the writing of Fragment B demonstrates that it should be placed below Fragment A. Five particular letters highlight this for us.<\/p>\n
Beth. The form of beth 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-1 and A-5. I propose that the curved-stem beth was employed in Fragment A because of the scribe\u2019s 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 portions of the stone, however, no such strain was exerted in writing. As such, the scribe was able to write vertexed-stem beths with ease.<\/p>\n
Daleth. The form of daleth in Fragment A was like an equilateral triangle with a small tail. The figure was fairly tall. It was also noticed how the horizontal stroke of the figure was written on different angles. This confirms that the scribe had some difficulty in writing the letter and was unable to achieve full consistency. The height of the figure testifies to the fact that the scribe was stretched considerably forward when writing. This made diagonal lines considerably steeper rather than flatter. On Fragment B, though, we see that the horizontal stroke is consistently written sloping up to the right. The form is also noticeably shorter or flatter than the form in Fragment A. This indicates that the scribe had greater freedom of movement with his arm. Thus there was no heaviness of hand to make the figure longer and to mark the horizontal stroke inconsistently.<\/p>\n
Heh. The difference between the hehs was seen to be the distance between the \u2018bristles\u2019 as well as their lengths. The form in Fragment A puts a noticeable gap between the \u2018bristles\u2019, 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 \u2018bristles\u2019 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 \u2018bristles\u2019. The stretched posture also prevented him from making the \u2018bristles\u2019 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 \u2018bristles\u2019 of heh A-6 are closer together and are of comparable length.<\/p>\n
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\u2019s 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.<\/p>\n
Resh. The same factor that was seen for the letter daleth applied to the writing of the resh. That is, the \u2018head\u2019 of the resh in Fragment A is noticeably taller than the \u2018head\u2019 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.
\nThus, the palaeographical evidence suggests that Fragment B should be placed below Fragment A.<\/p>\n
Textual Considerations
\nThe 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. Therefore, we must test to see whether their reading of the text is sustainable.
\nNumerous scholars have raised objections to the publishers\u2019 text. Indeed, 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in Fragment A, Line 9. However, 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 connective text in Line 1, I shall start with their reconstruction of Line 2.<\/p>\n
Figure 5.3. Facsimile of the fragments integrating the arrangement proposed by Biran and Naveh (drawn by Ada Yardeni).<\/p>\n
Line 2: [ ]\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05e1\u05e7\u05af[\u00b7\u05e2\u05dc\u05d5\u05d4\u00b7\u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4\u00b7\u05d1\u05d0-[-\u2026]. Between Fragment A and Fragment B, Biran and Naveh propose inserting six letters and two word dividers: [\u00b7\u05e2\u05dc\u05d5\u05d4\u00b7\u05d1\u05d4]. There is, however, barely enough room for six letters, let alone an extra two word dividers. The space required for one 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 other\u2014a suggestion which cannot really be sustained with any credibility. Thus, the reconstructed text of Biran and Naveh is physically impossible here.
\nBiran and Naveh offer no reconstruction for the end of the line.<\/p>\n
Line 3: \u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05d4\u05da\u00b7\u05d0\u05dc[\u00b7\u05d0\u05d1\u05d4\u05d5]\u05d4\u00b7\u05d5\u05d9\u05e2\u05dc\u00b7\u05de\u05dc\u05db\u05d9[\u05e9]. Between Fragment A and Fragment B, Biran and Naveh here propose inserting four letters and a word divider\u2014[\u00b7\u05d0\u05d1\u05d4\u05d5]\u2014as well as the full figure of heh which is only partially preserved in Fragment B. There is adequate room for this reconstruction.
\nAt the end of the line, Biran and Naveh propose adding one letter, a \u0161in, after restoring the yodh partially broken at the edge of Fragment B. This implies that the left edge of the whole inscription was immediately after this \u0161in, for the scribe ran out of room to complete the remaining three letters of the word \u05d9\u05e9\u05e8\u05d0\u05dc. These extra three letters were subsequently written at the beginning of the next line.<\/p>\n
Line 4: \u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd\u00b7\u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05d1\u05d9[\u00b7\u05d5]\u05d9\u05af\u05d4\u05de\u05dc\u05da\u00b7\u05d4\u05d3\u05d3[\u00b7]\u05d0[\u05d9\u05ea\u05d9\u00b7]. As Biran and Naveh have arranged the fragments, they propose inserting one letter and a word divider\u2014[\u00b7\u05d5]\u2014in the gap between Fragment A and Fragment B. There is adequate room for this reconstruction.
\nAt the end of the line, Biran and Naveh restore the \u2019aleph along the broken edge of Fragment B and add three letters and a word divider: [\u05d9\u05ea\u05d9\u00b7]. However, this runs into difficulty on the premise that only one letter could be added to the end of Line 3. This being the case, there is insufficient room to accommodate all the letters that Biran and Naveh wish to restore here at the end of Line 4. There is room enough for only two letters and a word divider as opposed to the three letters which Biran and Naveh propose. If indeed the left edge of the inscription permitted the addition of all three letters and the word divider, then we would expect to have had an extra letter at the end of Line 3\u2014namely, the resh in \u05d9\u05e9\u05e8\u05d0\u05dc. This letter, however, is at the beginning of Line 4 according to Biran and Naveh. Thus, there is some discrepancy as to the spacing at the end of Lines 4 and 5.<\/p>\n
Line 5: \u05d0\u05e0\u05d4\u00b7\u05d5\u05d9\u05d4\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05e7\u05d3\u05de\u05d9[\u00b7\u05d5]\u05d0\u05e4\u05e7\u00b7\u05de\u05df\u05af\u00b7\u05e9\u05af\u05d1\u05e2\u05af[\u05ea\u00b7—]. In this line, Biran and Naveh propose the insertion of one letter and one word divider\u2014[\u00b7\u05d5]. There is adequate room for this reconstruction.
\nAt the end of the line, Biran and Naveh propose the restoration of the partially damaged \u2018ayin 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 insertion of a taw and a word divider after the restored \u2018ayin, 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\u2019s left edge. If we allow Line 4 to dictate the position of the inscription\u2019s left edge, we have room enough for two letters and one word divider. However, Line 6 begins with a yodh, 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\u2019s left edge from Line 3 or Line 4.<\/p>\n
Line 6: \u05d9\u00b7\u05de\u05dc\u05db\u05d9\u00b7\u05d5\u05d0\u05e7\u05ea\u05dc\u00b7\u05de\u05af\u05dc\u05af[\u05db\u05df\u00b7\u05e9\u05d1]\u05e2\u05af\u05df\u05af\u00b7\u05d0\u05af\u05e1\u05e8\u05d9\u00b7\u05d0[\u05dc\u05e4\u05d9\u00b7\u05e8]. Upon restoring the damaged mem and lamed at the end of the line in Fragment A, Biran and Naveh insert four letters and a word divider\u2014[\u05db\u05df\u00b7\u05e9\u05d1]. After this, they restore an \u2018ayin as the first letter only partially visible on Fragment B. There is certainly room for this construction, but there seems to be too much room. There is extra space the size of one word divider or half a letter. Thus, for Biran\u2019s and Naveh\u2019s reconstruction, we must assume the letters were spread out more in this gap than in the rest of the line.
\nAt the end of the line, Biran and Naveh add four letters and a word divider\u2014[\u05dc\u05e4\u05d9\u00b7\u05e8]. If we allow Line 3 to dictate the position of the inscription\u2019s left edge, then we find that there is room enough for three letters. The word divider would have to sit outside the margin. If we allow Line 4 to dictate the position of the inscription\u2019s left edge, then the word divider can be accommodated comfortably. However, no matter which line we allow to dictate the position of the left edge, there is certainly no room for the resh which Biran and Naveh\u2019s reconstruction requires at the end of the line. This resh is the first letter of the word \u05e8\u05db\u05d1, the last two letters of which are written at the beginning of Line 7.
\nOn the other hand, if we allow Line 6 to dictate the position of the inscription\u2019s left edge, so that the four letters and one word divider which Biran and Naveh add are accommodated, then we have discrepancies with the end of Line 3. There, we would have more than enough room to accommodate at least the resh of \u05d9\u05e9\u05e8\u05d0\u05dc 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.
\nThus, 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.<\/p>\n
Line 7: \u05db\u05d1\u00b7\u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9\u00b7[\u05e7\u05ea\u05dc\u05ea\u00b7\u05d9\u05d4\u05d5]\u05e8\u05dd\u00b7\u05d1\u05e8\u00b7[\u05d0\u05d7\u05d0\u05d1\u00b7]. In the gap between Fragment A and Fragment B, Biran and Naveh require ten letters and three word dividers to be inserted\u2014[\u05e7\u05ea\u05dc\u05ea\u00b7\u05d0\u05d9\u05ea\u00b7\u05d9\u05d4\u05d5]. There is, however, only room enough for nine letters and three word dividers. This means that the waw of the theophoric element\u2014\u05d9\u05d4\u05d5, which Biran and Naveh add to the letter \u05e8\u05dd\u2014to form the name of Jehoram must be dropped. This, however, is not possible since the waw in the theophoric element\u2014\u05d9\u05d4\u05d5 is never dropped. The only other alternative is to suggest the apocopated form,\u2014\u05d9\u05d5. This yields the name \u2018Joram\u2019, an attested alternative to Jehoram which is used in the biblical text.
\nAt the end of the line, Biran and Naveh insert the personal name, Ahab. That is, they add four letters and one word divider\u2014[\u05d0\u05d7\u05d0\u05d1\u00b7]. If Line 3 dictates the left edge of the inscription, then we do not have enough room to include the word divider. If Line 4 dictates the left edge, then the word divider fits perfectly. If Line 6 dictates the left edge of the inscription, then there would be a blank space the size of one letter at the end of this line. Since Biran and Naveh\u2019s reading of the text requires Line 6 to dictate the position of the inscription\u2019s left edge, then we must propose this extra space at the end of Line 7, as well as the extra blank space at the end of Lines 3 and 4.<\/p>\n
Line 8: \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc\u00b7\u05d5\u05e7\u05ea\u05dc\u05af[\u05ea\u00b7\u05d0\u05d9\u05ea\u00b7\u05d0\u05d7\u05d6]\u05d9\u05d4\u05d5\u00b7\u05d1\u05e8[\u00b7\u05d9\u05d4\u05d5\u05e8\u05dd\u00b7\u05de\u05dc]. Biran and Naveh rightly restore the visible but damaged letters at the end of Fragment A, Line 8 as \u05e7\u05ea\u05dc. To this they add a taw to form the word \u05d5\u05e7\u05ea\u05dc\u05ea (\u2018and I killed\u2019). All in all, between Fragment A and Fragment B they insert seven letters and two word dividers. There is, however, only enough room for six letters and two word dividers. No apocopation can be suggested for their reading. Even cramming the letters closer to each other does not leave enough room for the extra letter required by their reading.
\nAs such, their reading here is unsustainable. Yet, it is one of the pivots of their reading that the name of Ahaziah (i.e. A\u1e25azyahu) can be restored here. To maintain the placement of Fragment B to the left of Fragment A, either Biran and Naveh must abandon their reading of Ahaziah\u2019s name here, or else arrange the fragments differently. This automatically destroys their entire reconstruction of the text, for even moving Fragment B slightly more to the left in order to accommodate the name of Ahaziah here creates problems for the letters to be inserted in previous lines. Thus, Biran and Naveh\u2019s reading cannot be maintained as it is physically impossible.
\nAt the end of the line, we have even greater difficulties accommodating Biran and Naveh\u2019s reading. They propose reading the kaph at the beginning of Line 9 as the last letter in the word \u05de\u05dc\u05da. As such, they require the addition of seven letters and two word dividers after the extant end of Line 8 in Fragment B. However, there is no way all these letters can be accommodated. If Line 3 dictates the position of the inscription\u2019s left edge, then there is only enough room for four letters and one word divider. If Line 4 dictates the left edge of the inscription, then only five letters and one word divider can be accommodated. If Line 7 dictates the position of the left edge, then there is enough room for only five letters and two word dividers. Even allowing for the apocopated form of the name Jehoram\u2014Joram\u2014we still do not have enough room for all the characters required by for the reading proposed by Biran and Naveh.
\nAlternatively, if we allow all the seven letters and two word dividers to be accommodated so that the position of the inscription\u2019s left edge is governed by this line, we have even greater discrepancies at the end of previous lines.
\nThis analysis shows that the arrangement of the fragments proposed by Biran and Naveh is unsustainable. We see that their reconstructed text inserted between the Fragments in Lines 2 and 8 is physically impossible due to lack of space. In addition, we see huge discrepancies with the placement of letters at the end of the lines. Yet, it is the placement of these letters, both between the fragments and at the end of the lines, on which Biran and Naveh\u2019s reading hangs.<\/p>\n
Orthographical Considerations
\nA major inconsistency also exists within the orthography of the text suggested 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 \u05de\u05dc\u05db\u05d9\u05e9\u05e8\u05d0\u05dc (\u2018the king of Israel\u2019). They note the possibility of such a compound without a word divider by appealing to a similar compound construction, \u05de\u05dc\u05db\u05d2\u05d1\u05dc (\u2018king of Byblos\u2019), in the tenth-century BCE Phoenician inscriptions of A\u1e25iram and Ye\u1e25imilk from Phoenicia. However, it is a regular orthographic feature of these particular inscriptions to write construct chains as one compound word without any word dividers. This is not the case with the Tel Dan fragments. In Line A8, we see the construction \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc with a word divider separating the two nouns of the construct chain. Unlike the fragmentary nature of Biran and Naveh\u2019s reading for the end of Line B3, this construction is complete. Therefore, it must form the basis for understanding the orthography of the inscription. This being the case, it is clearly evidence that compound orthography is not employed in the Tel Dan Inscription. Rather, the nouns of construct chains are clearly delineated by the use of word divider. Further evidence for this is seen in the constructions \u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9 (Line A7), \u05d0\u05af\u05e1\u05e8\u05d9\u00b7\u05d0[\u2026] (Line B6) and \u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05dc\u05af\u05d1\u05d9[\u2026] (Line A4).
\nNeedless to say, this has implications for our understanding of the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Line A9). However, it is abundantly clear from the unambiguous cases that word dividers are used in construct chains. In order for Biran and Naveh\u2019s reading to stand, we must posit that there is an inconsistent use of word dividers in two identical constructions. This is highly improbable given the evidence of other construct chains in the Tel Dan Inscription. Thus, we must consider Biran and Naveh\u2019s reading as erroneous and their arrangement of the fragments as unlikely.<\/p>\n
Restored Letters
\nBiran and Naveh\u2019s reading is further undermined by the restoration of certain letters on Fragment A, which were made in the epigraphical analysis.<\/p>\n
Line A4. Over the lacuna on Line A4, the remains of a lamed were identified. This means that Biran and Naveh\u2019s reading of this word as \u05d0\u05d1\u05d9 (\u2018my father\u2019) cannot be upheld. Rather, the word is to be restored as \u05d0\u05dc\u05af\u05d1\u05d9[\u2026] and regarded as fragmentary. As a result, Biran and Naveh\u2019s 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.<\/p>\n
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 \u05de\u05dc\u05db\u05df (\u2018kings\u2019). 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.
\nHowever, 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 \u05de\u05dc\u05db\u05df. Thus, Biran and Naveh\u2019s 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.<\/p>\n
A New Arrangement<\/p>\n
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.
\nA particular comment by Biran and Naveh is also pertinent to this proposal 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 reference to the author\u2019s father in Fragment A. 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\u2019s reference to his father. Furthermore, the text of Fragment B does not support its placement above Fragment A for there are no references to the author\u2019s father. Also, the content is more appropriate in the latter portions of a royal memorial inscription than in the earlier portions prior to references of the author\u2019s father.
\nGalil, however, opts for placing Fragment B above Fragment A. Unfortunately, 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\u2019s study lacks a detailed epigraphical examination to inform him of the line trends of each fragment and what they entail.
\nThese facts, coupled with the unsustainability of Biran and Naveh\u2019s reconstructed text, leave us with only one possibility for arranging the fragments\u2014namely, 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.
\nIn 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 inscription. 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\u2019s length (about 45 cm) of the bottom edge of the inscription. However, we have no clue as to where in the stele\u2019s width we should place Fragment B. That is, the longitude of Fragment B is unknown. 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.
\nThis arrangement sees Fragment B placed approximately 32 cm below Fragment A. Thus, there is a considerable gap between the fragments (see Fig. 5.4) implying contextual distance between the texts of the respective fragments. It is with this arrangement of the fragments that we must proceed to conduct a textual analysis.<\/p>\n
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 centimeters).<\/p>\n
Chapter 6<\/p>\n
TEXTUAL ANALYSIS<\/p>\n
Introductory Remarks<\/p>\n
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.<\/p>\n
Reference and Sigla
\nThe system of line reference used here will be a compound label, consisting of the fragment label and the line number within that fragment. For example, the second line in Fragment A is referred to as \u2018Line A2\u2019, and the fourth line in Fragment B is referred to as \u2018Line B4\u2019.
\nWe do not know whether word dividers were employed at the end of a line. For the sake of clarity, though, I have chosen to include them where I perceive that the last word of a line was complete and did not continue onto the next line. Letters that are mostly damaged, yet still restorable, are marked by a small circle above the letter (X\u030a). Lacunae are marked with brackets [ ]. Within some of the lacunae I have reconstructed particular words. Some of these reconstructions are less certain than others. I have ventured, however, to include only what may be plausibly suggested given the current evidence. I have avoided reconstructions founded solely on speculation or creativity. The reconstructions are individually dealt with in the commentary that follows the transcriptions.<\/p>\n
The Text<\/p>\n
Fragment A: Transcription<\/p>\n
(A1) [……………….]\u05d7\u05af\u05e9\u05e8\u00b7\u05e2[\u05dc\u00b7………………………………………]
\n(A2) [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05e1\u05e7\u05af [\u00b7………………………………………….\u00b7]
\n(A3) \u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05d4\u05da\u00b7\u05d0\u05dc[……………………………………….\u00b7\u05d1\u05db\u05dc\u00b7\u05d0]
\n(A4) \u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd\u00b7\u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05dc\u05af\u05d1\u05d9[\u05ea\u05d0\u05dc………………………………………\u00b7]
\n(A5) \u05d0\u05e0\u05d4\u00b7\u05d5\u05d9\u05d4\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05e7\u05d3\u05de\u05d9\u00b7\u0307[………………………………………..\u00b7\u05d9\u05d5\u05de]
\n(A6) \u05d9\u00b7\u05de\u05dc\u05db\u05d9\u00b7\u05d5\u05d0\u05e7\u05ea\u05dc\u00b7\u05de\u05af\u05dc\u05af[\u05da\u00b7]\u05d5\u05af[……………………………………\u05d0\u05dc\u05e4\u05d9\u00b7\u05e8]
\n(A7) \u05db\u05d1\u00b7\u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9\u00b7[…………………………………………..\u05d9\u05d4\u05d5\u05d0\u05d7\u05d6\u00b7]
\n(A8) \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc\u00b7\u05d5\u05e7\u05af\u05ea\u05af\u05dc\u05af[\u05ea\u05d4\u00b7…………………………………………\u05de\u05dc]
\n(A9) \u05da\u00b7\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3\u00b7\u05d5\u05d0\u05e9\u05af\u05dd\u05af\u00b7[……………………………………………….\u00b7\u05d0]
\n(A10) \u05d9\u05ea\u00b7\u05d0\u05e8\u05e7\u00b7\u05d4\u05dd\u00b7\u05dc[……………………………………………………..]
\n(A11) \u05d0\u05d7\u05e8\u05df\u00b7\u05d5\u05dc\u05d4[?………………………………………………\u05d9\u05d4\u05d5\u05d0\u05e9\u00b7\u05de]
\n(A12) \u05dc\u05da\u00b7\u05e2\u05dc\u00b7\u05d9\u05e9[\u05e8\u05d0\u05dc……………………………………………….\u05e9\u05de\u05ea\u00b7]
\n(A13) \u05de\u05e6\u05af\u05e8\u00b7\u05e2\u05dc[\u00b7\u05e9\u05de\u05e8\u05df ……………………………………………………]<\/p>\n
Fragment A: Translation<\/p>\n
(Al) […………] you will rule ov[er …………………………………………………………..]
\n(A2) [and because of the p]iou[s act] s of my father, may [?] go up [……………..]
\n(A3) and my father will repose. May he go to [……………………………….at every]
\n(A4) ancient [h]earth on ground of El-Bay[tel………………………………………..am]
\n(A5) I, so Hadad would go before me […………………………………………the day-]
\n(A6) -s of my reign, and I would slay a kin[g] and [……………..thousands of cha-]
\n(A7) -riots and thousands of horsemen[…………………………………………………….]
\n(A8) the king of Israel, and [I] killed [him…………………………………………….kin-]
\n(A9) -g of Bayt-Dawid. And [the] name of [……………………………………………..]
\n(A10) their land to[…………………………………………………………………………………]
\n(A11) another and to [………………………………………………………………Jehoash r-]
\n(A12) -eigned over ls[rael…………………………………………………………………I laid]
\n(A13) siege to [Samaria………………………………………………………………………….]<\/p>\n
Fragment B: Transcription<\/p>\n
(B1) […………………………\u00b7]\u05d5\u05d2\u05d6\u05e8\u05af[…………………………………………..]
\n(B2) […………………. \u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4\u00b7\u05d1\u05d0[?…………………………………….]
\n(B3) […………………]\u05d4\u05af\u00b7\u05d5\u05d9\u05e2\u05dc\u00b7\u05de\u05dc\u05db\u05d9\u05af[\u00b7\u05d4\u05d3\u05d3…………………………………….]
\n(B4) [………………]?\u00b7\u05af\u05d4\u05de\u05dc\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05d0\u05af[\u05d9\u05ea\u05d9…………………………………….]
\n(B5) [……………..\u05de]\u05d0\u05e4\u05e7\u00b7\u05de\u05df\u05af\u00b7\u05e9\u05af\u05d1\u05e2\u05af[\u00b7\u05de\u05dc\u05db\u05df…………………………………….]
\n(B6) [……………………..]\u05e2\u05af\u05df\u05af\u00b7\u05d0\u05af\u05e1\u05e8\u05d9\u00b7\u05d0[\u05e9……………………………………..]
\n(B7) [………………………..]\u05e8\u05dd\u00b7\u05d1\u05e8\u00b7[…………………………………………]
\n(B8) [……………….\u05d0\u05de\u05e6] \u05d9\u05af\u05d4\u05d5\u00b7\u05d1\u05e8[\u00b7\u05d9\u05d5\u05d0\u05e9………………………………………]<\/p>\n
Fragment B: Translation<\/p>\n
(B1) [……………………] and [?] cut [………………………………………….]
\n(B2) [……………..in] his [f]ighting against A[?………………………………]
\n(B3) [……………..]?. But my king, [Hadad,] would come [……………..]
\n(B4) [………………….] Hadad made m[e] king [……………………………]
\n(B5) [……………… b]raver than seven [kings………………………………..]
\n(B6) [………………]ty captured m[en…………………………………………..]
\n(B7) [………………….?[ram son of […………………………………………….]
\n(B8) [………………Amaz]iah son of [Joash……………………………………]<\/p>\n
Commentary on Fragment A<\/p>\n
Line A1
\nThe 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.
\n[\u2026]\u05ea\u05af\u05e9\u05e8\u00b7\u05e2[\u2026]. Biran and Naveh misunderstood the initial markings here as a mem. Closer epigraphical observations reveal the markings to be two distinct letters, a taw followed by a \u0161in. The difficulty in interpretation here is that we do not know if the letters \u05ea\u05af\u05e9\u05e8 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.
\nNo noun \u05ea\u05e9\u05e8 is attested in Northwest Semitic languages. There is a suggested emendation to a \u1e24atrean text (49.3) which replaces \u05d3\u05e0\u05e9\u05e8\u05d0 with \u05ea\u05e9\u05e8. However, no meaning is offered for \u05ea\u05e9\u05e8. Alternatively, we could see \u05ea\u05e9\u05e8 as the end of a personal name with the final two letters (\u05e9\u05e8) being the noun \u2018prince\u2019 or \u2018potentate\u2019. A less likely option is to view this as a defective spelling for the month \u2018Tishri\u2019. In order to maintain this, however, 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.
\nWe find more meaningful options for this word if we treat it as a verb. First, due to the taw preceding the \u0161in, we do not have any metathesis. This means we must be dealing with either a middle-weak root (\u221a\u05e9\u05d5\u05e8 or \u221a\u05e9\u05d9\u05e8), or else a verb in a verbal stem other than Ithpa\u2018al. Second, we have no root \u05ea\u05e9\u05e8 attested in Northwest Semitic languages. This leaves us with the option that the taw is a verbal preformative. The possibilities here for the verbal root of the element \u05e9\u05e8 are:<\/p>\n
1. \u221a\u05e9\u05c1\u05e8\u05e8\u2014\u2018to be steadfast\u2019. This would yield a Pe\u2018al Impft 3.fem.sg.\/2.masc.sg.
\n2. \u221a\u05e9\u05c2\u05e8\u05e8\u2014\u2018to rule, be prince\u2019. This also would be a Pe\u2018al Impft 3.fem.sg.\/2.masc.sg.<\/p>\n
In both these instances, the gemination of the verbal root is masked in the verbal form of the Pe\u2018al stem. Since both these verb options are human actions, we may reasonably dismiss the possibility that the subject was feminine. Females are rarely, if ever, mentioned in ancient Northwest Semitic inscriptions. The human subjects of such monumental inscriptions are almost exclusively males. The only real exception to this would be a reference to a female deity. However, from a statistical point of view, we should view this verb as masculine.
\nThe attractiveness of the second option is that the \u2018ayin, which begins the next word, may be understood as the first letter of the preposition \u05e2\u05dc (\u2018over\u2019). This yields a good idiomatic expression, \u05ea\u05af\u05e9\u05e8\u00b7 \u05e2\u05dc (\u2018you will rule over\u2019). Similar expressions placing this verbal root with the preposition \u05e2\u05dc can be seen in the Hebrew text of Num. 16:13 and Judg. 9:22. Although the fragmentary nature of the Tel Dan text does not allow us to identify the subject of this verb, we can logically suggest the subject is the author of the text, especially since the subject is likely to be masculine. This requires us to view the verb as part of direct speech, perhaps spoken by the author\u2019s royal predecessor or patron deity.<\/p>\n
Line A2
\n[\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7. Again, the first portions of the line are missing or damaged because of the stone\u2019s 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 a yodh immediately before this. The preceding 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 [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af.
\nThe noun \u05d7\u05e1\u05d3 is attested in the singular in later Jewish Aramaic from Palestine with the meaning \u2018piety\u2019. Although this would appear to be under the influence of Biblical Hebrew, where it is far more common, an adjectival form \u05d7\u05e1\u05d9\u05d3 is known from Punic epigraphic texts. Contextual analysis shows that it implies an act of kindness, often unwarranted, motivated by the subject\u2019s commitment to the recipient of the action.
\n[\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9. The addition of a yodh to the root \u05d7\u05e1\u05d3 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 1.com.sg. pronominal suffix (\u2018my piety\u2019), or a plural construct (\u2018pious acts of\u2019). The following words make the first option unlikely since subjects commonly follow verbs (see below). Therefore, we may view [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af as being in construct with the following word, \u05d0\u05d1\u05d9. The construct chain yielded, then, is [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9 (\u2018and because of the pious acts of my father\u2019).
\nThe speaker is clearly the author of the inscription. The impetus for adding a prefixed preposition \u05d1 to the restored word \u05d7\u05e1\u05d3\u05d9 comes from a similar expression in other royal inscriptions. In Panammu II, line 19, we find the expression \u05d1\u05e6\u05d3\u05e7\u00b7\u05d0\u05d1\u05d9\u00b7\u05d5\u05d1\u05e6\u05d3\u05e7\u05d9 (\u2018because of my father\u2019s righteousness and my righteousness\u2019). 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\u2019s ascension to the throne. Similarly, in lines 1\u20132 of Panammu II, we find the expression \u05d1\u05e6\u05d3\u05e7\u00b7\u05d0\u05d1\u05d4 (\u2018because of his father\u2019s righteousness\u2019). The stele of Si\u2019-gabbari from Nerab also contains an expression with the word \u05d1\u05e6\u05d3\u05e7\u05ea\u05d9 (\u2018because of my righteousness\u2019). These expressions require the prefixed preposition \u05d1, understood with causative sense, before the noun \u05e6\u05d3\u05e7 (\u2018righteousness\u2019). I propose a similar reading here, though with the noun \u05d7\u05e1\u05d3 (\u2018piety\u2019).
\nThe addition of a waw conjunction is motivated by the physical limitations of Fragment A. After adding the prefixed preposition \u05d1, we see there is room enough for only one letter or a word divider at the beginning of the line. Since the phenomenon of starting a line with a word divider is unattested in Fragment A, nor in any other monumental lapidary inscription, we must opt for the addition of a letter. The only letter that suffices to be added before the prefixed preposition \u05d1 is the waw conjunction.
\n[\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05e1\u05e7\u05af[\u2026]. The expression [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9 must be understood as a construct chain for syntactical reasons. The word following \u05d0\u05d1\u05d9 is the verb \u05d9\u05e1\u05e7\u05af. Since it is conventional syntax for verbs to precede their specified subject when they are placed next to each other, \u05d0\u05d1\u05d9 should probably not be seen as the direct subject of \u05d9\u05e1\u05e7\u05af. Biran and Naveh, however, did understand \u05d0\u05d1\u05d9 as the direct subject of \u05d9\u05e1\u05e7\u05af. This was probably due to the circumstantial text created by the fragment\u2019s breakage. That is, since Line A2 is mostly destroyed before the word \u05d0\u05d1\u05d9, and since the line is no longer extant past the qoph of \u05d9\u05e1\u05e7\u05af, it is easy to associate these two words together.
\nHowever, there is no real syntactical basis for doing this. The syntactical construction of a verb preceding its subject is common to all Northwest Semitic inscriptions and far exceeds the occurrences when the subject directly precedes the verb.
\nThe only evidence that may be brought against this comes from the Mesha Stele and the two Nerab Stelae. In lines 2\u20133 of the Mesha Stele, we read \u05d0\u05d1\u05d9\u00b7\u05de\u05dc\u05da\u00b7\u05e2\u05dc\u00b7\u05de\u05d0\u05d1\u00b7\u05e9\u05dc\u05e9\u05df\u00b7\u05e9\u05ea\u00b7\u05d5\u05d0\u05e0\u05da\u00b7\u05de\u05dc\u05db\u05ea\u05d9\u00b7\u05d0\u05d7\u05e8\u00b7\u05d0\u05d1\u05d9 (\u2018My father reigned over Moab 30 years, and I have reigned after my father\u2019). In both these clauses, we see the subject preceding the verb. In the stele of Sin-zer-ibni from Nerab, we read in lines 9\u201310, \u05e9\u05d4\u05e8\u00b7\u05d5\u05e9\u05de\u05e9\u00b7\u05d5\u05e0\u05db\u05dc\u00b7\u05d5\u05e0\u05e9\u05da\u00b7\u05d9\u05e1\u05d7\u05d5\u00b7\u05e9\u05de\u05da\u00b7\u05d5\u05d0\u05e9\u05e8\u05da\u00b7\u05de\u05df\u00b7\u05d7\u05d9\u05df (\u2018May Sahar and Shamash and Nikkal and Nusk drag your name and your place out of life!\u2019). Here we see the four expressed subjects of the verb \u05d9\u05e1\u05d7\u05d5 placed before the verb itself. Similarly, in the stele of Si\u2019-gabbari from Nerab, we read in lines 9\u201310, \u00b7\u05e9\u05d4\u05e8\u00b7\u05d5\u05e0\u05db\u05dc\u00b7\u05d5\u05e0\u05e9\u05da\u00b7\u05d9\u05d4\u05d1\u05d0\u05e9\u05d5\u00b7\u05de\u05de\u05ea\u05ea\u05d4\u00b7\u05d5\u05d0\u05d7\u05e8\u05ea\u05d4\u00b7\u05ea\u05d0\u05de\u05d3 (\u2018May Sahar and Nikkal and Nusk make his dying odious, and may his posterity perish!\u2019). Again, we see two clauses in which the expressed subjects precede their corresponding verbs.
\nThe syntax of the two clauses from the Mesha Stele, however, are of a particular construction\u2014namely, X + qa\u1e6dal. 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.
\nThe syntax of the examples from Nerab appears quite compelling in making us connect \u05d0\u05d1\u05d9 in Line A2 with the following verb \u05d9\u05e1\u05e7\u05af. 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 corresponding verbs of preformative verbal conjugations in the Tel Dan Inscription. The preformative verbal form \u05d5\u05d9\u05d4\u05da 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, \u05d4\u05d3\u05d3. We therefore have the syntactical construction in which the preformative verbal form precedes its subject. It must also be mentioned that the verbal forms from the Nerab Stelae are jussives. Although the verb \u05d9\u05e1\u05e7\u05af 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 Sefir\u00e9 Treaties. 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 (\u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9). Thus, although it is possible to associate \u05d0\u05d1\u05d9 with \u05d9\u05e1\u05e7\u05af, the possibility is remote. On contextual and syntactical grounds, therefore, I propose connecting \u05d0\u05d1\u05d9 in Line A2 not with the following verb, \u05d9\u05e1\u05e7\u05af, but with the preceding word, [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af.
\nAll the references from other texts cited by Biran and Naveh in support of their reading to place \u05d0\u05d1\u05d9 with \u05d9\u05e1\u05e7\u05af actually count against them. Two references from the Sefir\u00e9 Treaties are cited (I A.5 and I C.3\u20134) in which the verb \u05d9\u05e1\u05e7 or \u05d9\u05e1\u05e7\u05df appears in a subordinate relative clause. 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. The Biblical Hebrew reference from 1 Sam. 25:13 has no expressed 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 \u05d0\u05d1\u05d9 as the expressed subject of \u05d9\u05e1\u05e7\u05af.
\nFurthermore, since Line A2 is broken off at the qoph of \u05d9\u05e1\u05e7\u05af, we cannot even be sure whether this verb is singular or plural. This adds further uncertainty to a direct connection between \u05d0\u05d1\u05d9 and \u05d9\u05e1\u05e7\u05af. Reasonably, therefore, we must connect \u05d0\u05d1\u05d9 with the preceding word, [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af, yielding a construct expression, [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af\u00b7\u05d0\u05d1\u05d9 (\u2018and because of the pious acts of my father\u2019).
\nTwo possible implications derive from this expression. Either the author of the inscription mentioned his own piety before that of his father (in contrast 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\u2019s 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 mention his own piety, we may have an expression similar to lines 1\u20132 of Panammu II. There we read \u05d0\u05d1\u05d9\u00b7\u05e4\u05e0\u05de\u05d5\u00b7\u05d1\u05e6\u05d3\u05e7\u00b7\u05d0\u05d1\u05d4\u00b7\u05e4\u05dc\u05d8\u05d5\u05d4\u00b7\u05d0\u05dc\u05d4\u00b7\u05d9\u05d0\u05d3\u05d9\u00b7\u05de\u05df\u00b7\u05e9\u05d7\u05ea\u05d4 (\u2018My father Panammu, because of his father\u2019s righteousness, did the gods of Ya\u2019di deliver him from destruction\u2019). In the case of the Tel Dan Inscription, the author would not be referring to the pious acts of his father\u2019s father, but simply to those of his father.
\nReferences to an author\u2019s 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\u2019s father at the end of the Panammu II Inscription must be classed differently since the author\u2019s 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 appropriate 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 isolation or in clear distinction to those of his father if his father were the main referent of the text.
\n\u05d9\u05e1\u05e7\u05af[\u2026]. This verb confirms that the language of the inscription is Aramaic since the root of this verb, \u221a\u05e1\u05dc\u05e7, is attested only in Aramaic. Furthermore, the verbal form allows us to classify the language more specifically as Old Aramaic. Later strands of Aramaic retain the lamed of \u221a\u05e1\u05dc\u05e7 in the imperfect of the Pe\u2018al conjugation, and only in the Haph\u2018el or Aph\u2018el conjugations does the lamed elide. The lack of a heh or \u2019aleph preformative in the current form would seem to preclude the verb from being in one of these conjugations. The dialect of the Deir \u2018Alla texts (if it can even be classed as Aramaic), and that of the Sam\u2019alian texts from Zenjirli, are the only known dialects in which the preformative heh of the Haph\u2018el conjugation is lost in the imperfect. The close similarity between the dialect of the Tel Dan Inscription and the Sefir\u00e9 Treaties suggests that it is likely the heh preformative of the Haph\u2018el conjugation was retained in the Tel Dan Inscription, as it is in the Sefir\u00e9 Treaties. There is no prefixed mem to suggest an infinitive form.
\nThis leaves the Pe\u2018al or Pa\u2018el conjugations as options for this verb. Elsewhere, the verb is only attested once in the Pa\u2018el stem as a perfect in Egyptian Aramaic texts, but the meaning is ambiguous. It could mean \u2018to settle (a debt)\u2019 or \u2018to reach (an agreement)\u2019. This implies a causative sense. However, the causative of \u221a\u05e1\u05dc\u05e7 is most commonly attested with a Haph\u2018el or Aph\u2018el conjugation. There is only one occurrence of the root in the Pa\u2018el stem. As such, we are justified in seeing this verb in the Pe\u2018al conjugation, meaning simply \u2018to go up\u2019. As mentioned previously, the breakage at the edge of the fragment means we cannot discern from the epigraphic evidence whether this verb is singular or plural.
\nMost notable is the fact that this verb is in a preformative verbal conjugation. Biran and Naveh understood this as a Pe\u2018al imperfect 3.masc.sg. of \u221a\u05e1\u05dc\u05e7, but with a past tense meaning. In evidence, they mention the fact that imperfects are often used with the force of past tense in biblical prose and also in other inscriptions, such as Mesha, Zakkur and the Deir \u2018Alla texts. However, the examples cited by Biran and Naveh are far from being trouble-free and require some discussion.
\nFirst, the Hebrew of biblical prose and the Moabite of the Mesha Stele may be isolated as clear examples of \u2018Canaanite\u2019 idiom. The use of a preformative verbal conjugation in association with a prefixed waw consecutive is usual and frequent in Canaanite dialects. However, it is yet to be demonstrated that the Tel Dan Inscription does indeed represent a Canaanite dialect. On the contrary, the vocabulary confirms the language as Aramaic. Also, since we have no waw prefixed to this preformative verbal conjugation, or a preceding prepositional particle of any kind, it stands in seeming contrast to the norms of Canaanite dialects if we are to interpret \u05d9\u05e1\u05e7 with a past tense. It would require us to see \u05d9\u05e1\u05e7 as a preterite rather than an imperfect\u2014a phenomenon that is at home in Hebrew texts of a higher style, predominantly poetry rather than prose.
\nSecond, the other inscriptions to which Biran and Naveh find recourse in this instance are quite problematic. The plaster texts of Deir \u2018Alla represent a peculiar dialect and so will always present difficulties for interpreting other texts in their light. As Hackett has pointed out, the classification of the Deir \u2018Alla texts as a form of Aramaic has been unduly emphasized in the past. Rather, it is better to classify the Deir \u2018Alla texts as a type of South Canaanite dialect. The tense of the so-called \u2018consecutive imperfects\u2019 is not secure, especially given the fragmentary context in which they appear. It is also odd that these \u2018consecutive imperfects\u2019 appear only within the first seven lines of text. Similarly, the notion that the Old Aramaic inscription of Zakkur contains evidence for the usage of waw consecutive in Aramaic is also questionable. The few consecutive imperfects in Zakkur all appear within the same immediate context and may legitimately be understood as habitual imperfects rather than consecutives.
\nFurthermore, if there was any doubt about this, Muraoka has demonstrated that the usage of waw consecutive in Aramaic is highly improbable. Muraoka opts for the copulative use of prefixed waw in the Tel Dan Inscription, something he rather generously attributes the observation of to Biran and Naveh. By \u2018copulative\u2019, Muraoka means a form of prefixed waw where the phonetic structure is not identical with that represented by the waw consecutive of Biblical Hebrew (pata\u1e25 followed by gemination of the first preformative consonant). His reasoning for this is that, because there are no vowel points in ancient lapidary Aramaic texts, the phonetic structure of prefixed waws is ambiguous. Since all Aramaic texts employ perfect verbs to denote past tense, the alleged use of waw consecutive in the Aramaic texts of Zakkur and the Tel Dan Inscription is very rare and extremely inconsistent. Not only is it grammatically problematic, it is also contextually difficult to maintain. As such, it is highly unlikely that waw consecutive is employed at all in Aramaic.
\nAlthough the verb \u05d9\u05e1\u05e7\u05af in Line A2 is not prefixed with a waw, Muraoka\u2019s arguments, despite their circular nature, have a great deal of import for how we should interpret it. If \u05d9\u05e1\u05e7\u05af is to be interpreted as a past tense, it must be seen as a preterite rather than a consecutive imperfect. The distinction is very fine, but nonetheless important for the issue of waw consecutive usage in Aramaic. If Muraoka is correct in his hypothesis, then the preformative verbal conjugations in the Tel Dan Inscription may be seen as preterites (rather than imperfects) that take copulative waw. This would allow us to interpret the similar verbal conjugations in Zakkur and Deir \u2018Alla in the same way, without recourse to a rather inconsistent usage of waw consecutive in Aramaic.
\nSasson tries to posit a theory whereby this rather inconsistent usage of waw consecutive in Aramaic is rendered consistent. His argument is that waw consecutive, when used in prose, appears in the context of war or war-related events. As such, we should not be surprised to find waw consecutive used in Aramaic dialects when the context is directly or indirectly connected with the recounting of war. However, this theory fails to account for much of the usage of waw consecutive in Hebrew narrative that has nothing to do with war. For instance, it does not accommodate the usage of waw consecutive in the book of Ruth. It also implies that tales of war are necessarily the deciding factor in whether or not waw consecutive is used in prose. The corollary of this is that the war tale is the pinnacle or climax of a prose text whenever waw consecutive is employed. Yet, this does no justice to much of the biblical material. Also, Sasson\u2019s theory does not adequately explain why such a phenomenon is not employed in Phoenician, a much closer relative of Hebrew than Aramaic.
\nMuraoka\u2019s suggestion of copulative waw + preterite is quite attractive in explaining the usage of waw prefixed to preformative verbal conjugations in Aramaic. Certainly his arguments are more convincing than those of Sasson. Yet, the greatest weakness of Muraoka\u2019s theory is that he has not eliminated the seemingly random usage of these verbal forms. He has simply replaced the consecutive imperfect conjugation with the preterite conjugation. The question still remains why the text employs such preterite conjugations for past tense when it could be achieved by the far more commonly employed perfect conjugation. It is also unlikely that preterites are used at all in Aramaic. Therefore, we must ask whether the context even demands interpreting our preformative verbal conjugations in the Tel Dan Inscription with a past tense meaning.
\nIn order to answer this question, we must closely examine the context in which these particular verb forms appear. The Deir \u2018Alla texts are difficult to evaluate, partly because of the fragmentary nature of the texts, but also because of their peculiar dialect, which Hackett argues convincingly is closer to South Canaanite dialects than Aramaic. The seeming consecutive imperfects in the Zakkur Stele may legitimately be understood as habitual imperfects with durative quality (as opposed to simple imperfects with punctiliar quality). The context in which the verbs in question appear relates to how Zakkur beseeched his patron deity, Ba\u2018alshamayn, during the siege of Ha\u1e0frakh, and how Ba\u2018alshamayn responded with favourable oracles (Zakkur A.11\u201317). Five verbs (\u05d5\u05d0\u05e9\u05d0, \u05d5\u05d9\u05e2\u05e0\u05e0\u05d9, [\u05d5\u05d9\u05de\u05dc\u05dc], [\u05d5\u05d9\u05d0\u05de\u05e8] and \u05d5\u05d9\u05d0\u05de\u05e8) have commonly been regarded as consecutive imperfects in this context by virtue of the fact that they are preformative verbal conjugations prefixed with waw, yet with seemingly past tense meaning. However, since the simple perfect is used elsewhere in the Zakkur Stele to denote simple past tense, we must be sceptical about interpreting the relevant forms as consecutive imperfects. Of the five verbs, only three are actually extant on the preserved fragment of the stele. The other two verbs have been reconstructed. Nevertheless, it is logical to suggest that the siege of Ha\u1e0frakh, which occasioned Zakkur\u2019s appeal to Ba\u2018alshamayn, was a durative event. As such, we should view Zakkur\u2019s petition to Ba\u2018alshamayn as occurring throughout the duration of the siege. That is, Zakkur\u2019s petition was not a punctiliar action, but an ongoing, repetitive action. In the same way, we can view Ba\u2018alshamayn\u2019s favourable reponses through seers and messengers as an ongoing, repetitive action. It is pertinent that the deity\u2019s reponse came through more than one medium, implying that it was not a singular, isolated oracle or event which made the deity\u2019s attitude known. As such, the verbal forms that have been regarded as isolated instances of consecutive imperfects, or even preterites, in the Zakkur Stele should actually be regarded as habitual imperfects. This removes the need to explain an otherwise troublesome usage of consecutive imperfects in Old Aramaic.
\nWith regards to the Tel Dan Inscription, we must ask whether the contexts in which \u05d9\u05e1\u05e7\u05af and other preformative verbal conjugations appear demand a past tense meaning. This is partially hampered by the fragmentary context of the Tel Dan Inscription. Yet, we can still make meaningful observations.
\nMost verbs used in royal memorial inscriptions carry a past tense meaning. In Aramaic, this is expressed by the use of perfect verbs. Preformative verbal conjugations serve as either imperfects or jussives. It is difficult to understand \u05d9\u05e1\u05e7\u05af here in Line A2 as an imperfect (\u2018he will go up\u2019), but this is certainly not the case for a jussive (\u2018may he go up\u2019). The only contexts in which jussive verbs appear in royal memorial inscriptions are: (1) curse formulae (usually against potential vandals); (2) formulae for divine blessing; or (3) optative expressions of good will towards a deceased person in the afterlife. The context of Line A2 does not fit the first two options. The possibility of the third option, however, must be considered.
\nOptative statements expressing good will towards a deceased person in the afterlife are well attested in Northwest Semitic inscriptions. Such statements occur in Panammu I (lines 17, 21\u201322) and Panammu II (lines 17\u201318). Since they refer to a deceased king as the father of the living king, these expressions have import for the context of Line A2 (and Line A3). Also, these expressions invariably employ jussive verbal forms. The presence of both the word \u05d0\u05d1\u05d9, as well as the preformative verbal conjugation \u05d9\u05e1\u05e7\u05af, are compelling evidence that the context of Line A2 is an optative expression of good will towards the author\u2019s deceased father. Despite the fact that \u05d0\u05d1\u05d9 and \u05d9\u05e1\u05e7\u05af are in separate clauses, it is the context which is under scrutiny. In this regard, Line A4 must also be considered.<\/p>\n
Line A2 (end) and Line A4
\n\u05d0\u05d1\u05d9\u00b7\u05d9\u05e1\u05e7\u05af[\u2026\u00b7]\u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9\u00b7\u05d9\u05d4\u05da\u00b7\u05d0\u05dc[\u2026]. In Line A4 we have two preformative verbal conjugations (\u05d5\u05d9\u05e9\u05db\u05d1 and \u05d9\u05d4\u05da). Biran and Naveh understood both of these with past tense meaning. The crucial element is the verb \u05d5\u05d9\u05e9\u05db\u05d1. Here we find a preformative verbal conjugation with a prefixed waw, which Biran and Naveh understood as a consecutive. However, since the usage of waw consecutive is highly problematic in an Aramaic text, it is best to treat the suggestion here with a fair degree of scepticism. Also, the Zakkur Stele cannot really be used as evidence for waw consecutive in Aramaic. Therefore, I propose understanding this waw as a simple conjunction.
\nThis leaves us with the preformative verbal form, \u05d5\u05d9\u05e9\u05db\u05d1. We may understand this verb as either an imperfect or a jussive with its stated subject being the author\u2019s father (\u05d0\u05d1\u05d9). Since Biran and Naveh first understood this and other preformative verbal conjugations as having past tense meaning, all other scholars have erroneously followed the same line of interpretation. Biran and Naveh also misconstrued the nuance of \u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9, which they translated as \u2018and my father died\u2019. Yet, only in Rabbinic Aramaic does \u221a\u05e9\u05db\u05d1 actually denote the act of dying. Elsewhere, in reference to death, \u221a\u05e9\u05db\u05d1 describes the state of one after death. That is, it describes someone who is already dead, not someone at the moment of death. Therefore, we should not translate \u05d5\u05d9\u05e9\u05db\u05d1\u00b7\u05d0\u05d1\u05d9 as \u2018and my father died\u2019 in the way Biran and Naveh do. That would have to be expressed by \u221a\u05de\u05d5\u05ea.
\nCloser evaluation of the contextual usage of \u221a\u05e9\u05db\u05d1 in ancient Northwest Semitic languages gives us the following options for understanding the nuance here in Line A2:<\/p>\n
1. \u2018To have sexual relations\u2019.
\n2. \u2018To lie\/fall down\u2019.
\n3. \u2018To be in a lying position\u2019.
\n4. \u2018To sleep (as at night)\u2019.
\n5. \u2018To rest\u2019.
\n6. \u2018To repose in death\u2019.<\/p>\n
The only contextually meaningful nuances which may be seriously entertained for usage in a royal memorial inscription are the fifth and sixth options. Although the nuance of the second option may be taken as a reference to the author\u2019s father falling in battle, it would require a Haph\u2018el conjugation. The fifth option may provide a reference to the author\u2019s father resting comfortably in the afterlife. As such, it may be understood as a jussive in an optative expression (\u2018and may my father rest\u2019). The sixth option may be understood as either an imperfect or a jussive. As an imperfect, it may refer to the state of repose in which the author\u2019s father lies in death (\u2018and my father reposes\u2019). Alternatively, it may have a future sense as a reference to the ongoing repose of the author\u2019s father in the future (\u2018and my father will repose\u2019).
\n\u05d9\u05d4\u05da\u00b7\u05d0\u05dc[\u2026]. On the basis of context and the interpretation of preformative verbal conjugations in Aramaic, I propose understanding \u05d9\u05d4\u05da as a jussive form in an optative statement referring to the afterlife of the author\u2019s father. It is common to find the preposition \u05d0\u05dc after the verb of motion, \u221a\u05d4\u05d5\u05da. However, because of the fragmentary nature of the text, we do not know if there was a word divider immediately after the lamed of \u05d0\u05dc. As such, \u05d0\u05dc may just as plausibly be understood as the first letters of the subject of \u05d9\u05d4\u05da. In this regard, \u05d0\u05dc may be a theophoric element at the beginning of a personal name (\u2018May El-[\u2014] go\u2019). Alternatively, \u05d0\u05dc may be understood as the deity El (\u2018May El go up\u2019), or the deity El-Bethel (\u2018May El-Bethel go up\u2019). The original suggestion by Biran and Naveh was to read this as a reference to the author\u2019s father having died and gone \u2018to his house of eternity\u2019. However, \u05d9\u05d4\u05da should not be interpreted as a preterite, but as a jussive in an optative expression.<\/p>\n
Line A4 (end) and Line A4
\n[\u05d1\u05db\u05dc\u00b7\u05d0]\u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd\u00b7\u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05dc\u05af\u05d1\u05d9[\u2026]. The first three letters of Line A4 are the end of a word begun at the end of Line A4 (no longer extant). Biran and Naveh interpreted this as a reference to Israel ([\u05d9\u05e9]\u05e8\u05d0\u05dc) in the context of an Israelite king who formerly held sway over the land of the author\u2019s father, or who had simply entered the territory of the author\u2019s father. However, the restoration of the lamed over the lacuna makes a reference to the author\u2019s father here improbable. There is also nothing per se about the syntax of Line A4 which is definitive and which requires us to include all the elements within the same clause.
\nGarbini makes a notable protestation at Biran and Naveh\u2019s translation of \u05e7\u05d3\u05dd as an adverb (\u2018previously\u2019). He notes that the only times that the root \u05e7\u05d3\u05dd appears in an adverbial sense in Aramaic, it is as a plural noun, as a singular noun with a preceding preposition, or as an emphatic active participle. It is true that the adverbial sense is achieved most commonly with a plural noun (\u05e7\u05d3\u05de\u05df). However, when \u05e7\u05d3\u05dd is preceded by a preposition, it is not the preposition that necessarily gives it adverbial force. Rather, the adverbial sense is implied by \u05e7\u05d3\u05dd. Thus, in a Phoenician text from Cyprus, we see the expression, \u05d9\u05e8\u05d7\u00b7\u05de\u05d3\u00b7\u05d9\u05e8\u05d7\u00b7\u05e2\u05d3\u00b7\u05e2\u05dc\u05dd\u00b7\u05db\u05e7\u05d3\u05dd (\u2018month by month, forever as aforetime\u2019). In this way, \u05e7\u05d3\u05dd may be understood as a noun (\u2018aforetime\u2019) that has an implied adverbial sense. If we are permitted to import this understanding into Aramaic here, it would require \u05e7\u05d3\u05dd to be the stated circumstance of a new clause, thus separating it from the previous word. However, there are reasonable doubts as to whether the Phoenician expression can inform us of Aramaic idiom. Thus, Garbini\u2019s warning here is quite pertinent.
\nKnauf, De Pury and R\u00f6mer interpreted \u05e7\u05d3\u05dd as a perfect verb in the Pa \u2018el conjugation, yielding the expression, \u2018[The King of Is]rael had advanced into the land of my father\u2019. Although \u05e7\u05d3\u05dd is certainly attested as a verb in ancient Northwest Semitic texts, the difficulty here is that the authors have not restored the lamed over the lacuna towards the end of Line A4. As such, they read \u05d0\u05d1\u05d9 as the last extant word of Line A4, rather than the correct reading, \u05d0\u05dc\u0307\u05d1\u05d9.
\nIt is also unlikely that \u05e7\u05d3\u05dd should be read as a Pe\u2018al participle meaning \u2018chief\u2019. This nuance is attested in the Nabatean expression \u05e7\u05d3\u05de\u05d9\u00b7\u05e9\u05e8\u05db\u05ea\u05d4 (\u2018the chiefs of its corporation\u2019). However, it is unlikely that the author is here referring to himself as a \u2018chief\u2019 since this would be a lesser title than \u2018king\u2019. This raises the question of whether the author was indeed a king. Cryer raised this issue with reference to the word \u05de\u05dc\u05db\u05d9 in Line 6 as referring to the author\u2019s king. However, since this word should be read as a reference to the author\u2019s own \u2018reign\u2019, it confirms the royalty of the author. Therefore, \u2018chief\u2019 is an unlikely interpretation of \u05e7\u05d3\u05dd.
\nAlternatively, I propose understanding \u05e7\u05d3\u05dd as an adjective meaning \u2018ancient\u2019. In this way, it may be understood as attributive of a preceding masculine singular noun. The first three letters of Line A4 may then be understood as the end of the word [\u05d0]\u05e8\u05d0\u05dc. This enigmatic word appears in line 12 of the Mesha Stele in the context of booty captured by Mesha and brought before his patron deity, Kemosh. In this respect, the \u05d0\u05e8\u05d0\u05dc (or the \u05d0\u05e8\u05d0\u05dc\u00b7\u05d3\u05d5\u05d3\u05d4) appears to have had cultic significance. Ezekiel 43:15\u201316 describes an \u05d0\u05e8\u05d0\u05dc or \u05d4\u05e8\u05d0\u05dc as a type of four-homed altar par excellence, or an essential component of such an altar. As such, an \u05d0\u05e8\u05d0\u05dc is probably an altar hearth which may have been portable. The fire of such a hearth may have been symbolic of the deity\u2019s presence or attention, or may simply have been used to consume offerings made to the deity to whom the hearth was dedicated. This permits us to understand [\u05d0]\u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd is far from certain as a reference to an \u2018ancient altar hearth\u2019, mentioned in reference to sacred memorial rites for the author\u2019s deceased father. This fits the immediately preceding context very well. It also compares with Panammu I, lines 15\u201318 and 21\u201322, in which Panammu I entreats his successors to bless his memory and condition 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 [\u05d1\u05db\u05dc] before [\u05d0]\u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd is far from certain. However, it would not be inappropriate in the context.
\n\u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05dc\u05af\u05d1\u05d9[\u05ea\u05d0\u05dc \u2026]. Biran and Naveh understood \u05d0\u05e8\u05e7 as referring to the land of the author\u2019s father. However, since we restored a lamed over the lacuna in Line A4, this reading cannot stand. I propose connecting these two words to the preceding clause. We must consider \u05d0\u05dc\u05af\u05d1\u05d9[\u2026] as fragmentary since \u221a\u05d0\u05dc\u05d1 is unattested as a genuine root in Northwest Semitic languages. Although \u221a\u05d0\u05dc\u05d1 is attested as an orthographical variant of \u221a\u05d0\u05dc\u05e3 (\u2018to instruct, incite\u2019), we cannot make good contextual sense of this root here. In addition, we see the noun \u05d0\u05dc\u05e4\u05d9 in Line A7, making the interchange of peh with beth highly improbable in the Tel Dan Inscription. If the clause ended with \u05d1\u05d0\u05e8\u05e7, we might expect to find this noun in the emphatic state with the post-positional definite article (\u05d1\u05d0\u05e8\u05e7\u05d0). 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, 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 \u05d4\u05de\u05dc\u05da. 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 \u2019aleph. 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, similar to that in the Sefir\u00e9 Treaties. Therefore, the lack of a post-positional definite article here most probably indicates that \u05d1\u05d0\u05e8\u05e7 is in construct with the following word.
\nWith this understanding, I have restored \u05d0\u05dc\u05d1\u05d9[\u2026] to read \u05d0\u05dc\u05d1\u05d9\u05ea\u05d0\u05dc (\u2018El-Baytel\u2019). No other meaningful alternative for understanding this lexeme can be found. \u2018El-Baytel\u2019 is the hypostasis of the deity El, who came later to be known simply as \u2018Baytel\u2019 or \u2018Bethel\u2019. The name El-Baytel specifically refers to the deity El in the form of a ma\u1e63\u1e63ebah or sacred Bethel-stone at the city gate. In this form, El-Baytel watched over legal proceedings as a patron of wisdom and justice. The compound structure of the name compares with \u05d1\u05e2\u05dc\u05e9\u05de\u05d9\u05df (\u2018Ba\u2018alshamayn\u2019), \u05d0\u05e8\u05e7\u05e8\u05e9\u05e3 (\u2018\u2019Arq-Reshef\u2019), \u05e8\u05db\u05d1\u05d0\u05dc (\u2018Rakib-El\u2019), and most of the deities revered by the Judaean community of Elephantine\u2014\u05e2\u05e0\u05ea\u05d1\u05d9\u05ea\u05d0\u05dc (\u2018Anat-Bethel\u2019), \u05d0\u05e9\u05de\u05d1\u05d9\u05ea\u05d0\u05dc (\u2018Ishum-Bethel\u2019) and \u05d7\u05e8\u05de\u05d1\u05d9\u05ea\u05d0\u05dc (\u2018\u1e24erem-Bethel\u2019).
\nThe word \u05d0\u05e8\u05e7 here must be understood in the sense of a plot of \u2018ground\u2019, as opposed to \u2018earth\u2019, or \u2018land\u2019 in reference to a state entity. It refers to the ground on which an \u2018ancient altar hearth\u2019 ([\u05d0]\u05e8\u05d0\u05dc\u00b7\u05e7\u05d3\u05dd) stands, the ground being the sacred property of the deity El-Baytel. We may compare this with the biblical concept of \u2018holy ground\u2019 in which certain plots of ground were distinguished as special or consecrated to a deity. This concept is also borne out in temple architecture in which the temple sanctuary or main chamber (\u05e7\u05d3\u05e9) was a \u2018holy\u2019 place, and the room in which the deity resided was the \u2018holy of holies\u2019 (\u05e7\u05d3\u05e9 \u05e7\u05d3\u05e9\u05d9\u05dd).<\/p>\n
Line A5
\n\u05d0\u05e0\u05d4\u00b7\u05d5\u05d9\u05d4\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05e7\u05d3\u05de\u05d9\u00b7[\u2026]. Biran and Naveh understood the first word in Line A5 as the l.com.sg. independent personal pronoun, \u2018I\u2019. Originally, before the discovery of Fragment B, the context was understood as a statement concerning the author\u2019s humility, as in Zakkur I, line 2, \u05d0\u05d9\u05e9\u00b7\u05e2\u05e0\u05d4\u00b7\u05d0\u05e0\u05d4 (\u2018I am a humble man\u2019). After the discovery of Fragment B, this was altered to coincide with the arrangement of the fragments and the text proposed by Biran and Naveh. The new context was understood to be the author\u2019s legitimation of his own rulership through divine sanction, \u05d5\u05d9\u05d4\u05de\u05dc\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05d0\u05d9\u05ea\u05d9\u00b7\u05d0\u05e0\u05d4 (\u2018And Hadad made me king\u2019). This second reconstruction is now discredited because it was motivated by the arrangement of the fragments proposed by Biran and Naveh. This arrangement, we have seen, is practically impossible. It also appears to have been motivated by the syntactic disjunction inherent in understanding \u05d5\u05d9\u05d4\u05da as an example of waw consecutive. That is, \u05d0\u05e0\u05d4 was separated completely from what followed because the waw prefixed to \u05d9\u05d4\u05da was understood consecutively rather than conjunctively. Thus, the reading was subconsciously tinged with Hebraic hues.
\nSince it is unnecessary and unlikely that waw consecutive was a legitimate construction in Aramaic, we should understand this line differently from what Biran and Naveh propose. Thus, I view the waw prefixed to \u05d9\u05d4\u05da as conjunctive and the verb itself as an habitual imperfect rather than a consecutive imperfect. In this way, \u05d0\u05e0\u05d4 may be seen as the 1.com.sg. independent personal pronoun at the end of a clause, but still a part of the author\u2019s same train of thought that runs through the following clause. Due to the preceding context, the waw conjunction is understood with causative force leading to a result clause. Thus, in Lines A2\u20134, I understand the author to be remembering his deceased father\u2019s pious acts and wishing his father\u2019s memory to be perpetuated through cultic ritual. Now, in Line A5, I understand the author to be boasting of his own piety as king and the subsequent divine favour bestowed on him by Hadad.
\nBiran and Naveh\u2019s original suggestion that \u05d0\u05e0\u05d4 formed the end of the author\u2019s claim to humility is thus seen to be very plausible in the context. The end of Line A4 may have been similar or identical to the statement in Zakkur I, line 12: \u05d0\u05d9\u05e9\u00b7\u05e2\u05e0\u05d4\u00b7\u05d0\u05e0\u05d4 (\u2018A humble man am I\u2019). If the text preserves some kind of parallelism between the author and his father, the author may have claimed that he was a \u2018pious\u2019 man: \u05d0\u05d9\u05e9\u00b7\u05d7\u05e1\u05d3\u00b7\u05d0\u05e0\u05d4 (\u2018A pious man am I\u2019). Alternatively, the author may have claimed that he was a righteous king, as in Ye\u1e25awmilk, line 9: \u05db\u00b7\u05de\u05dc\u05da\u00b7\u05e6\u05d3\u05e7\u00b7\u05d4\u05d0 (\u2018for he is a righteous king\u2019); or Ye\u1e25awmilk, lines 6\u20137: \u05db\u05de\u05dc\u05da\u00b7\u05e6\u05d3\u05e7\u00b7\u05d5\u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8 (\u2018for [he] is a righteous king and an upright king\u2019).
\nThe statement that \u2018Hadad would go before me\u2019 compares with the biblical concept of the deity going before his people or his appointed leader as a guide or a conqueror. Primarily, it has a military connotation and 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 habitual imperfect \u05d5\u05d9\u05d4\u05da implies that the author was a regular military campaigner who, in his own estimation, enjoyed a good deal of success.<\/p>\n
Line A5 (end) and Line A6
\n[\u05d9\u05d5\u05de]\u05d9\u00b7\u05de\u05dc\u05db\u05d9. Originally, Biran and Naveh interpreted \u05de\u05dc\u05db\u05d9 as \u2018my king\u2019, inferring from it that the author was a vassal of a greater king. When Fragment B was published, Biran and Naveh reinterpreted it as \u2018my kingdom\u2019, a reading necessitated by their arrangement of the fragments. I propose understanding \u05de\u05dc\u05db\u05d9 as the abstract noun meaning \u2018my reign\u2019, with a 1.com.sg. pronominal suffix. This fits the context in which the author switches from discussing his deceased father to discussing his own exploits as king. The noun \u05de\u05dc\u05da meaning \u2018reign\u2019 or \u2018kingship\u2019 is attested in Sefir\u00e9 I B.6 as a singular construct noun, and probably also in I C.6 with a 1.com.sg. pronominal suffix, as here in Line A6. It is also attested in Ugaritic and Phoenician.
\nThe noun \u05de\u05dc\u05db\u05d9 is probably connected to the idea in the preceding line that Hadad would go before the author, bringing military success during the days of the author\u2019s reign. Obviously, some text is missing between the statement \u05d9\u05d4\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05e7\u05e8\u05de\u05d9 (\u2018And Hadad would go before me\u2019) and the phrase [\u05d9\u05d5\u05de]\u05d9\u00b7\u05de\u05dc\u05db\u05d9 (\u2018the days of my reign\u2019), but we may surmise that it has to do with the success granted to the author by Hadad and perhaps other gods. We may also expect that \u05d9\u05d5\u05de\u05d9 was immediately preceded by either \u05db\u05dc (\u2018all\u2019) or the prefixed preposition \u05d1 (\u2018during\u2019).
\n\u05d5\u05d0\u05e7\u05ea\u05dc\u00b7\u05de\u05af\u05dc\u05af[\u05da\u00b7]\u05d5\u05af[\u2026]. This phrase continues the author\u2019s claim that Hadad gave him military success by expounding on this idea. As such, \u05d0\u05e7\u05ea\u05dc should be understood with the same tense as \u05d9\u05d4\u05da in Line A5, that is, as an habitual imperfect (\u2018and I would slay\u2019), as opposed to the consecutive imperfect (\u2018and I slew\u2019) proposed by Biran and Naveh. Thus, the prefixed waw should be understood as the copulative rather than the consecutive. The retention of taw in \u05d0\u05e7\u05ea\u05dc, as opposed to an assimilation to emphatic \u1e6deth, is a common phonetic feature of Old Aramaic, also observable in the Sefir\u00e9 Treaties and the Sam\u2019alian texts.
\nThe word following \u05d5\u05d0\u05e7\u05ea\u05dc was misread by Biran and Naveh after the small remnant of a waw at the end of the line on the edge of Fragment A was mistaken for a nun. This yielded the erroneous reading \u05de\u05dc\u05db\u05df (\u2018kings\u2019). Although this word makes contextual sense, it was also partly shaped by the arrangement of the fragments proposed by Biran and Naveh. Epigraphically, there would be an inordinate gap between the kaph and nun. I propose that the waw at the edge of the fragment be interpreted as a waw prefixed to the beginning of another word. This leaves room enough for a kaph and a word divider after the restored lamed. Thus, our restored text reads \u05d5\u05d0\u05e7\u05ea\u05dc\u00b7\u05de\u05af\u05dc\u05af[\u05da\u00b7]\u05d5\u05af[\u2026] (\u2018and I would slay [a?] king and \u2026\u2019).
\nThe fact that \u05de\u05af\u05dc\u05af[\u05da] is not in the emphatic state supports the idea that the author is here making generalized claims about his military exploits. Also, the very fact that \u05de\u05af\u05dc\u05af[\u05da] is not preceded by the direct definite object marker, \u05d0\u05d9\u05ea, is pivotal. There is more than sufficient clear evidence throughout the inscription to suggest that, if \u05de\u05af\u05dc\u05af[\u05da] was a direct definite object of the verb \u05d5\u05d0\u05e7\u05ea\u05dc, it would have carried the direct definite object marker, \u05d0\u05d9\u05ea. Since this is not the case, we must understand \u05de\u05af\u05dc\u05af[\u05da] as an indefinite object within the clause. This being the case, it makes it even harder for those advocating the use of consecutive imperfects or preterites in the Tel Dan Inscription, and in Old Aramaic at large, to make their assertions. The syntax here severely undermines their case. We should bear in mind that the word after \u05de\u05af\u05dc\u05af[\u05da] begins with a waw. If we then take \u05d5\u05d0\u05e7\u05ea\u05dc as a consecutive imperfect or a preterite, as most do, then we must translate the phrase here as \u2018and I killed a king\u2019. Such usage of verbs makes for a very weak tone of voice, which hardly befits the genre of the inscription as a royal memorial stele. It is, therefore, clear that the use of preformative conjugations that allude to past events in the Tel Dan Inscription, and so also in Old Aramaic, must be understood as habitual imperfects. In this way, the wording and tone of the phrase here is seen to be assertive, as befits an inscription of this genre.
\nSince \u05de\u05af\u05dc\u05af[\u05da] 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\u2019s 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.<\/p>\n
Line A6 (end) and Line A7
\n[\u05d0\u05dc\u05e4\u05d9\u00b7\u05e8]\u05db\u05d1\u00b7\u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9\u00b7. This line continues the author\u2019s generalized claim of his military success, granted to him by Hadad. The desire for detail originally led Biran and Naveh to understand \u05d5\u05d0\u05dc\u05e4\u05d9 as a dual construct noun (\u2018two thousand\u2019). This was revised with the publication of Fragment B to an indefinite plural construct (\u2018thousands of\u2019). Given the generalizing tendency of the author at this point, I also propose interpreting \u05d5\u05d0\u05dc\u05e4\u05d9 as an indefinite plural. Muraoka\u2019s observation that early Aramaic inscriptions use cyphers to indicate numerals rather than actual lexemes is not applicable to the Tel Dan Inscription since it is evident from Lines B5 and B6 that numerals were spelled out with words rather than cyphers in the Tel Dan Inscription. The dual is also not usually employed where there would be ambiguity. Rather, numerals are employed, or else the dual is used of items occurring unambiguously in natural pairs or in clear contrast. These facts, along with the author\u2019s literary style, lead me to interpret \u05d5\u05d0\u05dc\u05e4\u05d9 as an indefinite plural noun in the construct state.
\nThus, the author is seen here to be making a generalized boast about the number of different armed units he defeated, destroyed, or captured during his reign. Alternatively, the author could be boasting here about the vastness of his own armed force. The former is most likely the case since it was customary style for a monarch to play up the strength of his enemy in order to enhance his own prestige as a result of victory. The previous generalization about killing kings also makes it likely that the references to armed units here are those of an enemy. The restoration of \u05d5\u05d0\u05dc\u05e4\u05d9 before [\u05e8]\u05db\u05d1 (\u2018chariots\u2019) is based on a parallelism with the phrase \u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9 (\u2018and thousands of horsemen\u2019).<\/p>\n
Line A7 (end) and Line A8
\n[\u05d9\u05d4\u05d5\u05d0\u05d7\u05d6\u00b7]\u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc. At this point, the author\u2019s generalized claims about his military success become more specific with a reference to the king of Israel. Since Hebrew, not Aramaic, was the language of Israel, this can hardly be a reference to the author himself being the king of Israel. Rather, this reference must be understood as applying to the author\u2019s enemy. Since the following word is a verb prefixed with waw, the words \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc must form the end of a clause. Therefore, I presume that the author\u2019s specific claims of military success began in the now non-extant portions of Line A7, perhaps with an expression stating that he went to war against the \u2018king of Israel\u2019. Thus, \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc would be seen as the indirect object of the clause. Alternatively, the \u2018king of Israel\u2019 may have been labelled as the belligerent party through a statement that the king of Israel went to war against the author. This would see \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc as the subject of the clause. In either possibility, the king of Israel was probably expressly named at the end of Line A7 since his successor seems to be expressly named in Line A11\u201312. For reasons pertaining to historical reconstruction, I have restored the Israelite king\u2019s name here as [\u05d9\u05d4\u05d5\u05d0\u05d7\u05d6] (\u2018Jehoahaz\u2019).
\n\u05d5\u05e7\u05af\u05ea\u05af\u05dc\u05af[\u05ea\u05d4]. Although the line breaks off after the lamed, the context allows us to restore a taw after it to maintain the author\u2019s speech in the first person. The restoration of a heh as a 3.masc.sg. object suffix is based on an understanding that the aforementioned \u2018king of Israel\u2019 is the object of the verb \u05e7\u05ea\u05dc. Once again, we see the retention of taw rather than an assimilation to emphatic \u1e6deth.
\nThis verb is very important for understanding the use of verbs in the entire inscription, as well as disproving the use of waw consecutive in Aramaic. The choice of an afformative verbal conjugation here signifies a past tense meaning. The fact that a preformative verbal conjugation was not chosen to convey this tense supports the case that these conjugations in the rest of the inscription are not to be considered as consecutive imperfects. Rather, the conveying of past tense meaning is fulfilled by the use of afformative verbal forms, whether the action is viewed perfectly, pluperfectly or consecutively. This is the normal convention in Aramaic and should, therefore, be applied here. We must translate this verb, then, with past-tense meaning as \u2018and [I] killed [him]\u2019.<\/p>\n
Line A9
\n\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. Without doubt, Line A9 has caused the greatest controversy arising from the Tel Dan Inscription. From Biran and Naveh\u2019s initial publication of Fragment A, debate has raged over the interpretation of the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. Biran and Naveh first posited the theory that it should be interpreted as \u2018House of David\u2019, referring to \u2018the dynastic name of the kingdom of Judah\u2019. A parallel was drawn to the Assyrian designations of various small states in the Levant, such as B\u00eet \u1e2aumri (\u2018House of Omri\u2019 = Israel), B\u00eet Agusi (\u2018House of Agusi\u2019 = Arpad) and B\u00eet \u1e2aaza\u2019ili (\u2018House of Hazael\u2019 = 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 A\u1e25ituv, Kallai, Andersen, Kitchen, Lemaire, Noll, Puech, Rainey, Rendsburg, Schniedewind, Tropper, Wesselius, and Yamada. On the other hand, there were those scholars who insisted that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 could not be interpreted as \u2018House of David\u2019, such as Cryer, Davies, Knauf, De Pury and R\u00f6mer, Lehmann and Reichel, Lemche and Thompson. It is obvious that this conundrum has yet to be resolved.
\nThe crux for interpreting the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 lies in the fact that there is no word divider between the seeming two parts, \u05d1\u05d9\u05ea and \u05d3\u05d5\u05d3. This suggests 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 confirmed by the fact that elsewhere in the Tel Dan Inscription, construct expressions are used to denote two or more concepts that are both individually 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, \u05d1\u05d0\u05e8\u05e7\u00b7\u05d0\u05dc\u05af\u05d1\u05d9[\u05ea\u05d0\u05dc] (\u2018on ground of El-Baytel\u2019). Similarly, in Line A8 we see the expression, \u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc (\u2018the king of Israel\u2019). In Line A7, we also have the expression, \u05d5\u05d0\u05dc\u05e4\u05d9\u00b7\u05e4\u05e8\u05e9 (\u2018and thousands of horsemen\u2019). We even see a surplus use of word dividers in Line A10 where a noun is demarcated from a pronominal suffix in the expression \u05d0\u05e8\u05e7\u00b7\u05d4\u05dd (\u2018their land\u2019). Thus, there can be no doubt that the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 denotes one idea rather than two separate ideas. Therefore, we must look to an interpretation that understands \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 as one essential entity. Unfortunately, none of the possible interpretations is completely devoid of problems.
\nOne of the more interesting hypotheses has been put forth by the German scholars, Lehmann and Reichel. With reference to the difficult text of Amos 8:14, they have supposed \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 to be a reference to a deity \u2018Dawd\u2019 or \u2018Dod\u2019. This understanding is achieved through a long-proposed amendment of Amos 8:14 in order to read \u05d5\u05d7\u05d9 \u05d3\u05d3\u05da \u05d1\u05d0\u05e8\u05be\u05e9\u05d1\u05e2 (\u2018by the life of your Dod, O Beersheba\u2019) rather than the Masoretic reading, \u05d5\u05d7\u05d9 \u05d3\u05e8\u05da \u05d1\u05d0\u05e8\u05be\u05e9\u05d1\u05e2 (\u2018by the life of the Way of Beersheba\u2019). By doing this, Lehmann and Reichel parallel the \u2018Dod of Beersheba\u2019 with the \u2018Ashima of Samaria\u2019 and the god of Dan also mentioned in Amos 8:14. A similar connection is then read into Line A9 whereby the lexeme \u05d5\u05d0\u05e9\u05dd, which occurs after \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3, is read as a reference to the goddess Ashima. Thus, Lehmann and Reichel here see a deity, Dod, in the guise of \u2018BaytDod\u2019, coupled with the deity Ashima, in the guise of \u2018Ashim\u2019, just as in Amos 8:14. Taking this understanding and transplanting it into the reading proposed by Knauf, De Pury and R\u00f6mer, they translate Line A9 as \u2018[I offered libat]ion to BaytDod and Ashim\u2019. In this way, Lehmann and Reichel compose the deity\u2019s name \u2018BaytDod\u2019 in the same way as other divine names known from Elephantine, such as Ishum-Bethel.
\nAttractive as this solution may sound, there are insurmountable difficulties precluding it. First, Lehmann and Reichel\u2019s interpretation necessitates a deity named \u2018Dod\u2019 to have been worshipped in the Levant during Iron II. Yet, it is highly unlikely that an independent deity named \u2018Dawd\u2019, \u2018Dod\u2019, or even \u2018BaytDod\u2019 was revered in the ancient Levant. A deity named \u2018Wadd\u2019 (\u2018Love\u2019) was worshipped in the central regions of Arabia before the advent of Islam, but a connection with an older Levantine deity named \u2018Dod\u2019 or \u2018Dawd\u2019 cannot be made. Rather, as Barstad and Becking have shown, \u05d3\u05d5\u05d3 is almost certainly to be regarded as a divine epithet applied to Yahweh and perhaps other deities known throughout the Levant, and probably northern Arabia also. Thus, if Lehmann and Reichel\u2019s theory is to have any applicability, \u2018Dod\u2019 must be understood as a divine epithet rather than an actual divine name.
\nNevertheless, there is no way this theory accounts for how the epithet (as opposed to the actual name) of an otherwise anonymous deity evolved from \u05d3\u05d5\u05d3 into the form \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. In the case of the deity El-Baytel there is a clear evolution of the name, which can be traced through a connection with sacred Bethel-stones. No such derivation can be traced for \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. So, although we may understand an epithet \u05d3\u05d5\u05d3 as \u2018beloved\u2019, we are at a loss to explain the form \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. Lehmann and Reichel no doubt identified this problem and proposed the possibility that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 be interpreted as a kind of cultic object. However, this destroys the parallel with the deity Ashim(a) connected to \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 with a waw conjunction.
\nFurthermore, the translation of Lehmann and Reichel, as well as that of Knauf, De Pury and R\u00f6mer, is discredited on numerous grammatical and syntactical grounds. First, these scholars suppose the use of consecutive imperfects in Aramaic. In order to maintain their reading, they interpret the single kaph at the beginning of Line A9 as the final letter of the word \u05d5\u05d0\u05e1\u05da (\u2018and I made libation\u2019). However, consecutive imperfects are highly unlikely in Aramaic. Rather, if the lexeme immediately preceding \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 was to be read as a verb with past tense, we would expect an afformative verbal form, \u05d5\u05e0\u05e1\u05db\u05ea, and hence a taw rather than a kaph at the beginning of Line A9. Otherwise, if this word were indeed a preformative verbal conjugation, we would have to interpret it as an habitual imperfect (\u2018and I would make libation\u2019). However, even this suggestion runs aground because the object of this verb would be the item poured out as a libation. With the given syntax, we would have to interpret \u2018BaytDod\u2019 as the object of the verb. That is, the syntax would require \u2018BaytDod\u2019 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. The prefixed preposition \u05dc is needed to understand \u2018BaytDod\u2019 as the deity to whom libation was made. Alternatively, the prefixed preposition \u05d1 is required to understand \u2018BaytDod\u2019 as a temple or shrine at which the libation was made. Since neither of these prepositions is present in the text, it is only with a great stretch of normal grammatical principles that we can avoid considering \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 as the direct object rather than the indirect object of a preceding verb. Even still, if we interpret \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 as a cultic object of some sort, as Lehmann and Reichel do, the waw conjunction prefixed to \u05d0\u05e9\u05dd (\u2018Ashim\u2019) 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 R\u00f6mer, have too many holes in them to be considered sustainable theories in regards to the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3.
\nThe suggestion that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 be translated as \u2018House of David\u2019 is also fraught with difficulties. The fact that the words \u05d1\u05d9\u05ea and \u05d3\u05d5\u05d3 have been combined into one composite lexeme means that the term was most probably not understood as a construct expression, \u2018House of David\u2019. 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, \u05d1\u05d9\u05ea\u00b7\u05d3\u05d5\u05d3. As such, the translation \u2018House of David\u2019 is impossible.
\nAlmost as a concession to this theory, however, is the fact that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 might refer to an actual state entity. We may consider \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 as a singular proper noun to be rendered \u2018Bayt-Dawid\u2019. 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 would not refer to the dynasty itself, but to the actual country ruled by that dynasty. That is, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 could be interpreted as a political term for a state entity. This is how such state labels of the type \u2018B\u00eet-PN\u2019 are used in Assyrian texts. Such usage is to be clearly distinguished from the proposal that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 be translated as a dynastic name, \u2018House of David\u2019. Nevertheless, this hypothesis by default incorporates an understanding that a \u2018House of David\u2019 at one time ruled a state entity.
\nHowever, the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 may just as reasonably be understood to refer to a geographical entity, rather than a political entity. That is, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 can be understood as a toponym referring to a town or district. In this way, the name would compare with composite toponyms such as \u05d1\u05d9\u05ea \u05e9\u05de\u05e9 (\u2018Beth Shemesh\u2019), \u05d1\u05d9\u05ea \u05dc\u05d7\u05dd (\u2018Beth Le\u1e25em\u2019), \u05d1\u05d9\u05ea \u05e9\u05d0\u05df (\u2018Beth She\u2019an\u2019), \u05d1\u05d9\u05ea \u05d0\u05dc (\u2018Bethel\u2019), \u05d1\u05d9\u05ea \u05d7\u05d5\u05e8\u05df (\u2018Beth \u1e24oron\u2019), \u05d1\u05d9\u05ea \u05d3\u05d2\u05d5\u05df (\u2018Beth Dagon\u2019), \u05d1\u05d9\u05ea \u05e8\u05d7\u05d1 (\u2018Beth Re\u1e25ob\u2019), \u05d1\u05d9\u05ea \u05de\u05e2\u05db\u05d4 (\u2018Beth Ma\u2018akah\u2019), and numerous others. It has been suggested by Thompson that within this interpretation, the element \u05d3\u05d5\u05d3 is the epithet of a deity, presumably Yahweh, meaning \u2018Beloved\u2019. This is certainly more plausible than the suggestion that \u2018Dawd\u2019 or \u2018Dod\u2019 was an independent Levantine deity.
\nThere are some minor difficulties with these two hypotheses, though. First, these composite names for either a state or a toponym are almost always written as two distinct lexemes. Such is the case in Aramaic, Hebrew and Moabite. There are, however, two known exceptions that occur, both of them with toponyms. The first is on Ostracon B from Tell Qasil\u00e9 which documents an import of 30 shekels of Ophir gold to \u2018Beth-\u1e24oron\u2019, which is written as a single lexeme, \u05d1\u05d9\u05ea\u05d7\u05e8\u05df. The second exception is a reference to a certain \u2018Bethel\u2019 of unknown location in Sefir\u00e9 I A.34, which is also transcribed as one word, \u05d1\u05d9\u05ea\u05d0\u05dc. Both these references are dated to the eighth century BCE. Thus, although most composite titles are written as two distinct words, there is certainly scope for seeing a toponym transcribed as a single lexeme.
\nThe second difficulty is that no town or district by the name \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 is known to us from the ancient record. This latter point, of course, does not mean that such a place did not exist in antiquity. To claim that no such place could have existed is to argue from silence. However, the lack of such a toponym in the ancient record does not inspire great confidence that such a place did in fact exist.
\nCryer called attention to an honorific stone inscription from Palmyra, in which the expression \u05d1\u05ea \u05d3\u05d5\u05d3\u05d0 appears. In the context, it refers to someone who was \u05e2\u05dc \u05d1\u05ea \u05d3\u05d5\u05d3\u05d0 (\u2018over the cookhouse\u2019). The expression may be rendered literally as \u2018the house of the cooking pot\u2019. Although this would present a parallel with the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Tel Dan Inscription, 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 \u2018cookhouse\u2019 in the Tel Dan Inscription would entail for both the context and general understanding of the inscription.
\nInevitably, we must interpret \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Tel Dan Inscription. Lines A7\u20138, as discussed above, refer to the author\u2019s 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3, the author refers to \u05d0\u05e8\u05e7\u00b7\u05d4\u05dd (\u2018their land\u2019). Presumably, one of the entities entailed in the 3.masc.pl. pronoun \u05d4\u05dd is the king of Israel. However, since the pronoun is plural, the author obviously envisages more than just the king of Israel here.
\nThis 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, \u2018their land\u2019 may refer to the land of the king of Israel and his armies. It is questionable, however, whether a king\u2019s armies would have been associated with him in the expression, \u2018their land\u2019, particularly since armies were traditionally seen as belonging to the king and, therefore, on a par with the land rather than with the king. This leads us to the second, more likely option\u2014namely, 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\u2019s enemy, and hence the author\u2019s usage of the 3.masc.pl. independent personal pronoun, \u05d4\u05dd in Line A10.
\nGiven that the reference to \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 appears in the line before this pronoun, we may understand \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 was the Aramaic equivalent of Hebrew \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019 or \u201cIr-Dawid\u2019). That is, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 is best understood as a reference to Jerusalem. We should, therefore, render \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in English as \u2018Bayt-Dawid\u2019, considering it a composite toponym transcribed as a single lexeme. We should then interpret the single kaph at the beginning of Line A9 as the last letter of the construct noun, [\u05de\u05dc]\u05da (\u2018[kin]g of\u2019). This implies that at the time the Tel Dan Inscription was written, Jerusalem was a city-state rather than the capital of a much wider regional state. These issues will be discussed in greater detail in Chapter 7.
\nWe may rightfully term this usage of \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 eponymous. Similar toponyms appear in the biblical literature in which a region is associated with one particular personage. We find various regions referred to by eponymous titles, understood as references to particular personages, such as \u05d1\u05d9\u05ea \u05d9\u05d5\u05e1\u05e3 (\u2018House of Joseph\u2019), \u05d1\u05d9\u05ea \u05d9\u05e9\u05e9\u05db\u05e8 (\u2018House of Issachar\u2019), \u05d1\u05d9\u05ea \u05d9\u05d4\u05d5\u05d3\u05d4 (\u2018House of Judah\u2019), \u05d1\u05d9\u05ea \u05e2\u05e9\u05d5 (\u2018House of Esau\u2019), and numerous others. Contrary to Rendsburg\u2019s assertions, these particular examples are not the result of Aramaic influence upon Judaean scribal schools, but simply labels carrying connotations of kinship or ethnicity. However, no such connotations are embedded in the label \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 of the Tel Dan Inscription. Rather, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 is closer in nuance to the biblical term \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019 or \u2018Ir-Dawid\u2019), and the Assyrian term mat \u1e2aumri (\u2018land of Omri\u2019 or \u2018Omri-land\u2019).
\nThis latter term should be clearly distinguished from the term B\u00eet \u1e2aumri (\u2018House of Omri\u2019), which is a reference to a political state or dynasty rather than a toponym. This distinction is highlighted by the fact that in Assyrian texts, no person is ever termed the king of a particular dynasty. That is, there is no formula \u0161ar B\u00eet-PN. Rather, such a concept was conveyed by the formula PN1 mar PN2. Thus, in the various texts dating to the reign of Shalmaneser III, Jehu of Israel is referred to as laua mar \u1e2aumri (literally, \u2018Jehu son of Omri\u2019). The connotation, however, is \u2018Jehu the Omrite\u2019. What this shows is that it is extremely unlikely that we have in the Tel Dan Inscription an expression [\u05de\u05dc]\u05da\u00b7\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 where \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 refers to a dynasty. Rather, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 must be a toponym. Thus, we can compare the expression [\u05de\u05dc]\u05da\u00b7\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Tel Dan Inscription to the expression \u05de\u05dc\u05da \u05d1\u05d9\u05ea\u05be\u05d0\u05dc in Josh. 12:16.
\nIn claiming this interpretation for \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3, however, there are a number of caveats to bear in mind. First, there is the already mentioned difficulty of interpreting \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 as a toponym. We expect it to have been transcribed as two separate words, rather than a single composite word. However, the evidence from Tell Qasil\u00e9 and Sefir\u00e9 demonstrates the rare but, nonetheless, attested practise of transcribing composite toponyms as a single lexeme. Second, the 3.masc.pl. pronoun, \u05d4\u05dd, may well refer only to the king of Israel and his armies as mentioned above. Although there are questions about this possibility, we cannot simply discount it. Such an interpretation, though, leaves the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 without interpretation. Yet, neither does this suggestion preclude \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 from being interpreted as a name for a city (or a state), for the pronoun \u05d4\u05dd in Line A10 might refer to the king of Israel and his armies without including a monarch of \u2018Bayt-Dawid\u2019 in its scope.
\nAlthough we cannot be perfectly certain that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 was intended as a reference to Jerusalem during a time when the city was called \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3, we can be confident that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 was indeed a toponym. The flow of the immediately surrounding context makes the proposed interpretation of \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 as a reference to Jerusalem most likely. However, since that context is still quite fragmentary, there will be a speck of doubt pending further recovery of other fragments. What is certain from the context is that the author is recounting a conflict in which the king of Israel played a role. The language of chariots, horsemen (Lines A7\u20138) and siege (Line A13) is undoubtedly that of war. The reference to the king of Israel is also unquestionable. It is clear that at least one other king was mentioned alongside the king of Israel. The most logical solution to this is to understand the second king as the ruler of a place called \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. As mentioned above, the context of the Tel Dan Inscription and other corroborated sources show us that Jerusalem is by far the most likely candidate for this place.
\nI cannot stress enough that \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 should be regarded as a toponym and not a reference to a Davidic dynasty. Although this label may have had an etymology going back to a Davidic dynasty, this is not how the author of the Tel Dan Inscription used it. Rather, the author was here referring to a geographical entity. My contention is that this geographical entity was Jerusalem. In the next chapter, this will be considered in more detail.
\n\u05d5\u05d0\u05e9\u05af\u05dd\u05af. This was mistakenly understood by Biran and Naveh as a consecutive imperfect meaning, \u2018and I put\u2019 or \u2018and I set\u2019. Rather, we should identify here the common noun for \u2018name\u2019 in Old Aramaic, prefixed with a waw conjunction to start a new clause. The prothetic \u2019aleph is a normal, though not universal feature, of this noun in Old Aramaic. It is well attested in the Sefir\u00e9 Treaties and Panammu I, though not in Zakkur. The suggestion of Lehmann and Reichel, mentioned above, was to understand this as the name of the deity Ashim(a). However, it was demonstrated that this was not a plausible theory on numerous grounds. Since Line A9 breaks off immediately after this word, it is impossible to specify the contents of this new clause. Since the author claims, in Line A8, to have killed the king of Israel, after mentioning him and his force, this here may be a reference to how the author dealt with the King of Bayt-Dawid by wiping out his name. This, however, is only speculation.<\/p>\n
Line A9 (end) and Line A10
\n[\u05d0]\u05d9\u05ea\u00b7\u05d0\u05e8\u05e7\u00b7\u05d4\u05dd\u00b7\u05dc[\u2026]. The first two letters on Line A10 are best understood as the final part of the accusative article, [\u05d0]\u05d9\u05ea. The \u2019aleph is restored in accordance with Old Aramaic. The oddity here is that the pronominal suffix, \u05d4\u05dd, has been separated from its relevant noun, \u05d0\u05e8\u05e7, by a word divider. This seeming overuse of word division led Biran and Naveh to mention the possibility of a reference here to \u2018the land of Ham\u2019. The only time such a land is ever mentioned in the ancient record is in Gen. 14:5. There are, however, numerous difficulties associated with this passage. It is far better to compare the separation of the pronominal suffix \u05d4\u05dd here in Line A10 with the same phenomenon in other Northwest Semitic inscriptions. Line 18 of the Mesha Stele has the expression \u05d5\u05d0\u05e1\u05d7\u05d1\u00b7\u05d4\u05dd (\u2018and I dragged them\u2019), Zakkur A.9 contains the expression \u05d5\u05de\u05d7\u05e0\u05d5\u05ea\u00b7\u05d4\u05dd (\u2018and their armies\u2019), while Bar Rakib B.7 has the expression, \u05e0\u05d1\u05e9\u05ea\u00b7\u05d4\u05dd (\u2018their souls\u2019). 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.
\nThe scope of \u05d4\u05dd 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 \u05d4\u05dd, it is interesting that the author uses only a singular noun, \u05d0\u05e8\u05e7 (\u2018land\u2019) in mentioning their territory. This may present a case for understanding the scope of \u05d4\u05dd as including only the king of Israel and his military units. More probably, the author simply regarded the territories 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 separate 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 borders 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.
\nThe 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 \u2018to\u2019 ruin. Biran and Naveh suggested that the word after the lamed here may have been \u05d9\u05e9\u05de\u05df (\u2018desolation\u2019). This word is employed in Sefir\u00e9 I A.32 in a curse formula. The suggestion is certainly plausible, but is still only speculation.<\/p>\n
Line A11
\n\u05d0\u05d7\u05e8\u05df. This adjective, \u2018another\u2019, is most probably used attributively of a preceding singular masculine noun written at the end of Line A10, but which is no longer extant. A substantive usage, however, cannot be ruled out. Since Fragment A grows considerably narrow in the last few lines, the exact context of this adjective is a matter of guesswork. The words \u05de\u05e6\u05af\u05e8\u00b7\u05e2\u05dc (\u2018siege against\u2019) in Line A13 clearly show that the author is still engaged in recounting armed encounters, but this does not aid us in understanding the precise context of Line A11. The suggestion of Biran and Naveh that the phrase might have been [\u05de\u05dc\u05da\u00b7]\u05d0\u05d7\u05e8\u05df (\u2018another king\u2019) is certainly plausible in the light of Line A12, which appears to introduce a new king of Israel. There is, however, no way of confirming this in the shrinking context. The same may be said for their other suggestion that this may have been a reference to \u2018another battle\u2019.
\n\u05d5\u05dc\u05d4[?\u2026]. The waw prefixed to this word introduces a new clause. Since the line breaks off so abruptly, however, we can only speculate as to what the whole word originally was. The lamed and heh are well preserved. Only a small fraction of the following letter is preserved. It was observed during epigraphical analysis that this letter can only be one of three letters, namely nun, peh or qoph. If we restore this letter to a nun, we may derive \u05dc\u05d4\u05df, understood either as the adversative coordinate conjunction \u2018however\u2019, or as the subordinate conjunction \u2018except\u2019. This, though, is unlikely to have been prefixed with another conjunction. Since none of the three possible letters form a verbal root with lamed and heh, we may reasonably deduce that the lamed is a prefixed preposition. This means that the word following could be either a noun or, more probably, an infinitive. This being the case, we may suggest the words \u05d5\u05dc\u05d4\u05e0\u05dd (\u2018and to remove\u2019) and \u05d5\u05dc\u05d4\u05e4\u05da (\u2018and to overturn\u2019) as possibilities. Alternatively, we may see the heh as indicative of the Haph\u2018el conjugation. Thus, we may also suggest \u05d5\u05dc\u05d4\u05e0\u05d6\u05e7 (\u2018and to cause harm\u2019), \u05d5\u05dc\u05d4\u05e0\u05d7\u05ea (\u2018and to bring down\u2019), \u05d5\u05dc\u05d4\u05e0\u05e4\u05e7 (\u2018and to bring out\u2019), \u05d5\u05dc\u05d4\u05e4\u05e7\u05d3 (\u2018and to give command of\u2019), \u05d5\u05dc\u05d4\u05e7\u05de\u05ea (\u2018and to raise\/ establish\u2019), and many others as suitable restorations. The numerous possibilities, though, serve to demonstrate that we cannot be sure what this word actually was.<\/p>\n
Line A11 (end) and Line A12
\n[\u05d9\u05d4\u05d5\u05d0\u05e9\u00b7\u05de]\u05dc\u05da\u00b7\u05e2\u05dc\u00b7\u05d9\u05e9[\u05e8\u05d0\u05dc \u2026]. The presence of the preposition \u05e2\u05dc means that the preceding word is most probably a verb (\u2018he reigned\u2019) rather than a noun (\u2018king\u2019). Although the preposition is the only complete word on this line, the other two words are easily restored because of the content in other lines. It is possible to translate [\u05de]\u05dc\u05da\u00b7\u05e2\u05dc\u00b7\u05d9\u05e9[\u05e8\u05d0\u05dc] as \u2018[was k]ing over Is[rael]\u2019, but this is less natural than reading \u2018[r]eigned over Is[rael]\u2019, which is to be preferred.
\nUnfortunately, the subject of the verb [\u05de]\u05dc\u05da is no longer extant, though it was presumably named at the end of Line A11. What is clear, however, is that if [\u05de]\u05dc\u05da 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. Thus, we should regard [\u05de]\u05dc\u05da 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 [\u05d9\u05d4\u05d5\u05d0\u05e9] on historical grounds.
\nA further point of interest is that the subject of the verb, [\u05de]\u05dc\u05da, actually precedes the verb. This may have been done to emphasize the name of the new Israelite king. Alternatively, the syntax might indicate that [\u05de]\u05dc\u05da is to be regarded as a participle. The same word order appears in the Mesha Stele, lines 4\u20135, with the phrase \u05e2\u05de\u05e8\u05d9\u00b7\u05de\u05dc\u05da\u00b7\u05d9\u05e9\u05e8\u05d0\u05dc. However, there also, there is some ambiguity as to whether \u05de\u05dc\u05da is a verbal form or a noun.<\/p>\n
Line A12 (end) and Line A13
\n[\u05e9\u05de\u05ea\u00b7]\u05de\u05e6\u05af\u05e8\u00b7\u05e2\u05dc[\u00b7\u05e9\u05de\u05e8\u05df]. Since the noun \u05de\u05e6\u05af\u05e8 (\u2018siege\u2019) is followed by the preposition \u05e2\u05dc (\u2018against\u2019), I presume it was preceded by a form of the verb \u05e9\u05d9\u05dd (\u2018to place\u2019) or \u05de\u05d7\u05d0 (\u2018to force\u2019), depending on the indirect object that was besieged. Both these verbs are used with the preposition \u05e2\u05dc to express the idea of laying a siege, the former to express siege against a location, and the latter to express siege against a person. Unfortunately, the fragmentary context precludes us from identifying what was besieged. However, historical considerations indicate that the author was probably the one who laid a siege here, probably during a campaign against \u2018their land\u2019 (Line A10)\u2014that is, the land of Israel and Bayt-Dawid. On these grounds, I suggest that it was a city that the author besieged, probably the city of Samaria.
\nHence, I have reconstructed the verb immediately preceding \u05de\u05e6\u05af\u05e8 as [\u05e9\u05de\u05ea] (\u2018I laid\u2019) and the indirect object of the verb as [\u05e9\u05de\u05e8\u05df] (\u2018Samaria\u2019).<\/p>\n
Commentary on Fragment B<\/p>\n
Line B1
\n[\u2026]\u05d5\u05d2\u05d6\u05e8\u05af[\u2026]. 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 \u221a\u05d2\u05d6\u05e8 (\u2018to cut\u2019). 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 context 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 letters without extra fragments of the stele.
\nThis 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 aggressor 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\u2019s own territory rather than in an opponent\u2019s 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 a defence of Gezer, written by Gezer\u2019s defender, ended up as building material all the way up in Dan. Also, we may survey the use of prefixed waw conjunctions in the other portions of text, including Fragment A, to aid us. This exercise shows that wows are prefixed to verb forms five times in Fragment A and once in Fragment B, as opposed to only two instances of waw prefixed to nouns. As a result, we are compelled to interpret \u05d5\u05d2\u05d6\u05e8\u05af as a verb rather than a noun. At the same time, we must realize that this conclusion is based on circumstantial evidence and so is not completely secure. We also cannot identify the person or number of the verb, though we may reasonably surmise the gender to be masculine. The verb \u221a\u05d2\u05d6\u05e8 is also common in Aramaic.<\/p>\n
Line B2
\n[\u2026 \u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4\u00b7\u05d1\u05d0[?\u2026]. The first extant word, despite being in fragmentary condition, is understood as a verb form from \u221a\u05dc\u05d7\u05dd (\u2018to fight\/battle\u2019), also known from Moabite and Hebrew. The remnant of a taw at the beginning of the line is the taw-infix of the Ithpe\u2018el conjugation, hence the restoration of heh before it. The taw-infix is also present in the Moabite forms from the Mesha Stele. However, the Moabite form employs a metathesis between the taw-infix and the first radical, lamed. This does not occur here in the Aramaic of the Tel Dan Inscription. Rather, the\u2014\u05d4\u05ea prefix is kept as a cohesive element of the conjugation. The restoration of the prefixed preposition \u05d1 is prompted by the presence of a 3.masc.sg. pronominal suffix, \u05d4\u2014(\u2018his\u2019). We are, therefore, dealing with an Ithpe\u2018el infinitive. A similar word, \u05d1\u05d4\u05dc\u05ea\u05d7\u05de\u05d4, which is identical except for the metathesis, is found on the Mesha Stele, line 19. The reference to a third person here (\u2018his fighting\u2019) is most likely to one of the author\u2019s enemies, probably intended as a defamatory statement of the enemy\u2019s unjustified aggression.
\nThe prefixed preposition \u05d1 on the following word should probably be understood in the sense of \u2018against\u2019. However, this is not absolutely certain. This preposition is idiomatically employed with verbs from \u221a\u05dc\u05d7\u05dd in the Mesha Stele to indicate the object fought against. Biran and Naveh, however, opt for translating this preposition as \u2018at\u2019, indicating the location of battle. 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 preposition as \u2018against\u2019. On the other hand, if it indicated a location, we could translate it as either \u2018against\u2019 or \u2018at\u2019. The evidence of the Mesha Stele suggests we translate it \u2018against\u2019. However, in the Mesha Stele we have a wider definite context allowing us to make this decision. The preposition \u05d1 is, however, used universally in this sense when preceded by a form of \u221a\u05dc\u05d7\u05dd. Therefore, we may suggest translating the preposition here as \u2018against\u2019, but we cannot make this a definite decision.
\nIn the earlier epigraphical analysis of Fragment B, it was observed that four letters stood as candidates for restoration of the letter following\u2014\u05d1\u05d0. These were beth, yodh, lamed and \u0161in. 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.<\/p>\n
Line B3
\n\u05d5\u05d9\u05e2\u05dc\u00b7\u05de\u05dc\u05db\u05d9\u05af[\u00b7\u05d4\u05d3\u05d3]. 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 \u221a\u05e2\u05dc\u05dc (\u2018to come, enter\u2019). The objections to this, which Becking raises, have more to do with Biran and Naveh\u2019s arrangement of the fragments and the difficulty in connecting Line B3 with the beginning of Line A4 than with translation of this verb. Becking\u2019s objections in regard to Biran and Naveh\u2019s arrangement and reading are appropriate, but the new text created by rearranging the fragments does not necessitate his interpretation that \u05d5\u05d9\u05e2\u05dc is derived from \u221a\u05e2\u05dc\u05d4 (\u2018to go up\u2019). The root, \u221a\u05e2\u05dc\u05d4, is common in Phoenician, Punic and Hebrew, but only enters Aramaic at a much later stage in the Targums, under the influence of Hebrew. Also, the concept of \u2018going up\u2019 is conveyed by \u221a\u05e1\u05dc\u05e7, which has already been seen in the Tel Dan Inscription in Line A2. Furthermore, seeing the lexeme \u05d5\u05d9\u05e2\u05dc as stemming from \u221a\u05e2\u05dc\u05d4 creates the false impression that this is a preterite verbal form because of the seeming elision of the \u05d4 from the root. However, as has been argued above, it is highly unlikely that preterite verb forms are used at all in conventional Old Aramaic. Therefore, we must interpret \u05d5\u05d9\u05e2\u05dc as a preformative verbal conjugation in the Pe\u2018al stem, derived from the geminate root \u221a\u05e2\u05dc\u05dc.
\nAs Becking notes, Biran and Naveh\u2019s understanding of the subject of this verb is not, however, viable. They saw the letters \u05de\u05dc\u05db\u05d9\u05af as forming a continuum with the beginning of Line A4, yielding the composite word \u05de\u05dc\u05db\u05d9\u05e9\u05e8\u05d0\u05dc (\u2018king of Israel\u2019). However, the context of Line A4 makes this suggestion highly questionable. Furthermore, the consistent, even excess use of word dividers means we must have grave doubts about the omission of a word divider here. Also, Biran and Naveh\u2019s interpretation was necessary in order to maintain the integrity of their arrangement, which is no longer sustainable. Therefore, we must interpret the letters \u05de\u05dc\u05db\u05d9\u05af differently.
\nTwo options are open to us. First, we may understand \u05de\u05dc\u05db\u05d9\u05af as the first four letters of a personal name, such as \u05de\u05dc\u05db\u05d9\u05af[\u05d4\u05d5]. Alternatively, we can interpret the yodh as a 1.com.sg. pronominal suffix of the noun \u05de\u05dc\u05da (\u2018king\u2019), yielding the word \u2018my king\u2019. I propose this latter option on the understanding that the author\u2019s reference to \u2018my king\u2019 is a reference to his patron god, Hadad. Thus, I suspect that the deity\u2019s name followed \u05de\u05dc\u05db\u05d9\u05af. This compares with the expression \u05de\u05e8\u05d0\u05d9 \u05e8\u05db\u05d1\u05d0\u05dc (\u2018my lord Rakib-el\u2019) in Bar Rakib A.5. The yodh of \u05de\u05dc\u05db\u05d9\u05af cannot be indicative of a plural construct noun or a plural noun with a pronominal suffix because the connection with the preceding singular verb, \u05d5\u05d9\u05e2\u05dc, would be lost. The idea of a deity being \u2018king\u2019 is attested in the Aqhat and Ba\u2018al cycles from Ugarit, in which it serves as an epithet for El. Yahweh is also deemed to be a king in Pss. 47:3; 95:3; Jer. 46:18; Mal. 1:14. It is also a dominant theme in 1 Samuel 8.
\nThe use of an habitual imperfect in regard to Hadad comes as no surprise, for the author implies Hadad\u2019s ongoing action on his behalf. The connotation is that the author had always been favoured by Hadad, shown by Hadad\u2019s continual demonstration of favourable actions towards him. We saw a similar connotation to the habitual imperfect used in reference to Hadad in Line A5, \u05d5\u05d9\u05d4\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05e7\u05d3\u05de\u05d9 (\u2018And Hadad would go before me\u2019). Here in Line B3, the author may be claiming that Hadad would always \u2018come\u2019 to his aid. This gloss is particularly suitable in the light of Line B2, which mentions an enemy\u2019s hostility. I suggest that the wow conjunction prefixed to \u05d9\u05e2\u05dc be understood with adversative force, \u2018but\u2019.<\/p>\n
Line B4
\n\u05d4\u05de\u05dc\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05d0\u05af[\u05d9\u05ea\u05d9]. We have understood the first extant word of Line B4 as a Haph\u2018el perfect of \u221a\u05de\u05dc\u05da. 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, \u2018Hadad made m[e] king\u2019, indicating a singular action in past tense. This compares with the speech of Ba\u2018alshamayn in Zakkur A.13, \u05d0\u05e0\u05d4\u00b7\u05d4\u05de\u05dc[\u05db\u05ea\u05da] (\u2018it was I who made you king\u2019). Here in the Tel Dan Inscription, though, the king is the speaker, rather than the deity.
\nCryer\u2019s suggestion that the heh of \u05d4\u05de\u05dc\u05da represents the definite article of Canaanite dialects should be dismissed, particularly since the rest of the Tel Dan Inscription reveals traits that clearly belong to conventional Old Aramaic. Cryer\u2019s 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.
\nThe lack of a waw conjunction prefixed to \u05d4\u05de\u05dc\u05da probably indicates that this lexeme occurs in mid-sentence rather than at the beginning of a fresh sentence. Exactly what preceded the word \u05d4\u05de\u05dc\u05da is uncertain. Our epigraphical analysis demonstrated that the last letter of the preceding word could only have been a gimel, heh or samekh. 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\u2019s enemies. However, the markings are too ambiguous to permit confident reconstruction of the sentence. What we may say, however, is that the words preceding \u05d4\u05de\u05dc\u05da were likely to be the reason given by the author for why Hadad made him king.
\nThe 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. However, since the object marker does not precede the verb, the emphasis is not as strong as it could have been.<\/p>\n
Line B5
\n[\u2026]\u05d0\u05e4\u05e7\u00b7\u05de\u05df\u05af\u00b7\u05e9\u05af\u05d1\u05e2\u05af[\u2026]. Due to the fragmentary context, it is extremely difficult to interpret this line. First, we do not know whether \u05d0\u05e4\u05e7 is a complete lexeme or whether other letters preceded it. Second, there are four different interpretations that may be offered, of which only the first two have previously been entertained.<\/p>\n
1. The first reading triggered by the fragments is to consider \u05d0\u05e4\u05e7 as the Pe\u2018al imperfect 1.com.sg. of \u221a\u05d0\u05e4\u05e7 (\u2018to go forth\u2019), as Biran and Naveh have done. This reading is prompted by the lines of the fragment\u2019s breakage, as well as the following lexeme, \u05de\u05df (\u2018from\u2019). Pursuing this line of interpretation, the context does not permit us to verify whether the form is an imperfect with future tense (\u2018I will go forth\u2019), a habitual imperfect (\u2018I would go forth\u2019), or even a cohortative (\u2018May I go forth\u2019). It is certainly not an example of a consecutive imperfect as Biran and Naveh proposed. 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, \u2018I would go forth from seven[\u2026]\u2019. 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 \u05e9\u05af\u05d1\u05e2\u05af[\u2026] in the context. Biran and Naveh\u2019s tentative suggestions, \u2018And I went forth (to war) outside the seven (?) [districts] of my kingdom\u2019, seem rather awkward and cumbersome, in spite the use of a consecutive imperfect here. It is also motivated by their proposed arrangement of the fragments. Nevertheless, we must keep the thrust of this suggestion because of uncertainty involved in the highly damaged context.
\n2. The second alternative, espoused primarily by Becking, is to understand \u05d0\u05e4\u05e7 as the name of a location, \u2018Apheq\u2019. At least four locations bearing the name Apheq are known in the region, namely the sites of Khirbet \u2019Afqa (near modern Beirut), Fiq (in the Golan), Tell el-Kurdaneh (10 km south of Akko), Ras el-\u2018Ain (at the headwaters of the Yarqon in the Sharon Plain) and probably a fifth but as yet undiscovered site in the Jezreel Valley. \u2018Apheq\u2019 could even be the toponym broken off at the end of Line B2 where a battle seems to have occurred. Indeed, a town called Apheq features prominently in the biblical account of the Syro-Ephraimite wars.
\n3. Another option is to regard \u05d0\u05e4\u05e7 as the noun meaning \u2018channel\u2019. This leaves us with the interesting proposition that the following word, \u05de\u05df, is the noun for \u2018water\u2019, spelled defectively without the yodh. This would yield the expression, \u2018a channel of water\u2019, presumably followed by a statement giving the length of the channel (\u05e9\u05af\u05d1\u05e2\u05af[\u00b7\u05de\u05d0\u05df\u00b7\u05d0\u05de\u05d4], \u2018seven hundred cubits\u2019)? The seemingly out-of-place reference to a water channel in what appears to be a war narrative cannot be dismissed because the context is so fragmentary as to defy precise definition.
\n4. Since we do not know what immediately preceded the \u2019aleph of \u05d0\u05e4\u05e7, we can surmise that \u05d0\u05e4\u05e7 was part of a larger lexeme\u2014namely, a Pa\u2018el participle, [\u05de]\u05d0\u05e4\u05e7, derived from \u221a\u05d0\u05e4\u05e7 (\u2018to be brave, determined\u2019). This word is probably also found in Sefir\u00e9 I B.29. The use of the following preposition, \u05de\u05df, is therefore comparative (\u2018than\u2019), giving us the expression, \u2018braver than seven[\u2026]\u2019. In light of Line B4, in which the author claims that Hadad made him king, this seems an appropriate boast for him to make. It allows us to continue the theme of military conflict encountered earlier while at the same time justifying the author\u2019s rule. Presumably, the author here compares his own might with that of seven opponents, presumably kings. However, since the context is so fragmentary, the suggestion cannot be confirmed. Nevertheless, this is my preferred reading, above the other three options.<\/p>\n
Line B6
\n[\u2026]\u05e2\u05af\u05df\u05af\u00b7\u05d0\u05af\u05e1\u05e8\u05d9\u00b7\u05d0[\u2026]. Biran and Naveh reconstructed the first extant letters on Line B6 as the end of the word [\u05e9\u05d1]\u05e2\u05af\u05df (\u201870\u2019). However, equally plausible are [\u05d0\u05e8\u05d1]\u05e2\u05af\u05df\u05af (\u201840\u2019) and [\u05ea\u05e9] \u05e2\u05af\u05df\u05af (\u201890\u2019), perhaps as the last part of a larger number. The difficulty with the following word, \u05d0\u05af\u05e1\u05e8\u05d9, 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, \u201870 kings who harnessed thousands of chariots and thousands of horsemen\u2019. Thus, their arrangement also inspired their interpretation of the single preserved \u2019aleph at the end of Line B6 as \u05d0[\u05dc\u05e4\u05d9] (\u2018thousands of\u2019). Yet, since Biran and Naveh\u2019s arrangement is dubious, we are not limited to this interpretation, plausible though it is in the new arrangement. If we take \u05d0\u05af\u05e1\u05e8\u05d9 as a passive, then we should translate the lexeme as \u2018captured\u2019 or \u2018prisoners of\u2019. This allows us to interpret the single \u2019aleph at the end of the line as perhaps the first letter of the plural noun, \u05d0[\u05e0\u05e9\u05df] (\u2018men\u2019), or the collective noun, \u05d0[\u05e9] (\u2018men\u2019), or even as the first letter of a toponym, such as \u05d0[\u05e8\u05dd] (\u2018Aram\u2019), or \u05d0[\u05d3\u05dd] (\u2018Edom\u2019). As a tentative interpretation, I suggest \u2018captured men\u2019, but with the realization that the context is too damaged to provide a definite interpretation.<\/p>\n
Line B7
\n[\u2026]\u05e8\u05dd\u00b7\u05d1\u05e8\u00b7[\u2026]. The final two lines of Fragment B have sparked nearly as much controversy as Line A9. Biran and Naveh unequivocally reconstructed this line as [\u05d9\u05d4\u05d5]\u05e8\u05dd\u00b7\u05d1\u05e8\u00b7[\u05d0\u05d7\u05d0\u05d1] (\u2018Jehoram, son of Ahab\u2019). The majority of objections raised to this interpretation were in regards to the historicizing that it entails. Indeed, it must be said that Biran and Naveh\u2019s interpretation represents a fairly subjective approach. Despite the tantalizing prospect of identifying a known historical figure here, we must refrain from jumping to conclusions without further consideration of the various issues.
\nInevitably, the crux of interpretation lies in the lexeme \u05d1\u05e8 and its possibilities for translation. The first instinct is to regard this as the common Aramaic word for \u2018son\u2019, just as Biran and Naveh did. Cryer\u2019s objection that Line B7 could not possibly be a reference to Jehoram, son of Ahab, does not bear much weight. Apart from verbally attacking Biran and Naveh\u2019s hypothesis, Cryer does not actually eliminate \u2018Jehoram, son of Ahab\u2019 as a possible reconstruction of Line B7. His suggestions that the lexeme \u05d1\u05e8 might be translated as \u2018cistern\u2019 or \u2018fortress\u2019 are rather weak. First, the Old Aramaic word for cistern preserves a diphthong which is written as \u05d1\u05d9\u05e8 in Sefir\u00e9 I B.34. Since it is evident from the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Line A9) that internal diphthongs are also preserved in the Tel Dan Inscription, it is unlikely that \u05d1\u05e8 in Line B7 is the noun meaning \u2018cistern\u2019. Second, the noun for fortress in Aramaic is feminine and is written as \u05d1\u05d9\u05e8\u05d4 in later strands of Aramaic. The word is unattested in Old Aramaic, but is nonetheless likely to have carried a feminine marker also.
\nApart from the noun \u2018son\u2019 as an interpretation of \u05d1\u05e8, we may also suggest the noun meaning \u2018outside\u2019. Although this noun is commonly attested in later strands of Aramaic, it presents no orthographical problems for Old Aramaic. It is also used in Biblical Aramaic and Hebrew with the connotation of \u2018open countryside\u2019. Alternatively, we can suggest the noun meaning \u2018grain\u2019, or the adjective meaning \u2018pure\u2019. The difficulty with these, however, is in interpreting the previous letters, \u05e8\u05dd, which are most certainly the end of a larger lexeme. One possibility is to see them as the end of a noun, such as [\u05db]\u05e8\u05dd (\u2018vineyard\u2019) or [\u05d7]\u05e8\u05dd (\u2018banned object\u2019), or perhaps the toponym, [\u05d0]\u05e8\u05dd (\u2018Aram\u2019). Or, we could see the final mem as perhaps being a 3.masc.pl. pronominal suffix. Yet, although the context here is extremely damaged, we still have enormous difficulties in understanding the syntax. As a result, I propose staying with the interpretation, \u2018son\u2019. This is further motivated by the fact that the author is almost certainly talking about people in Line B6 and is likely to have continued this onto Line B7.
\nFollowing this line of interpretation, it is unlikely that we have here a reference to someone whose name began with the element\u2014\u05d1\u05e8. The presence of a word divider immediately after \u05d1\u05e8 makes this improbable, especially since it was a general convention for such names to be written as single lexemes. The only known exception is the name \u05d1\u05df\u00b7\u05e2\u05d6\u05e8 (\u2018Ben-\u2018azzur\u2019) on Lachish Ostracon 19. However, the orthography of the Tel Dan Inscription suggests that such a name would have been transcribed as a single lexeme.
\nIt is also unlikely that \u05d1\u05e8 is an independent absolute noun, for both the meaning and syntax would then be impossible to decipher. This may simply be due to the highly damaged context, and indeed one must leave scope for such possibilities. Nevertheless, seeing \u05d1\u05e8 as an absolute noun requires us to interpret it as indefinite since no emphatic marker or definite article is attached to it. This would occasion a strange expression in such a monumental lapidary inscription. Therefore, I propose viewing the noun as being in a construct relationship with the following lexeme that, unfortunately, is no longer extant. All these observations mean that we must consider \u05e8\u05dd as the last part of a personal name, followed by the person\u2019s patronymic.
\nOur problem now lies in reconstructing this name. The element \u05e8\u05dd\u2014is well attested in the Northwest Semitic onomasticon. It may be preceded by a theophoric element, as in the names \u05d9\u05d4\u05d5\u05e8\u05dd (\u2018Jehoram\u2019), \u05d4\u05d3\u05d5\u05e8\u05dd (\u2018Hadoram\u2019), and \u05d0\u05dc\u05e8\u05dd (\u2018\u2019Ilram\u2019). It is also attested with various other elements, as in the names \u05d0\u05d1\u05e8\u05dd (\u2018\u2019Abiram\u2019), \u05d0\u05d3\u05e0\u05d9\u05e8\u05dd (\u2018\u2019Adoniram\u2019), \u05d0\u05d3\u05e0\u05dc\u05e8\u05dd (\u2018\u2019Adonleram\u2019), \u05d0\u05d7\u05e8\u05dd (\u2018\u2019A\u1e25iram\u2019), and \u05e2\u05de\u05e8\u05dd (\u2018Amram\u2019). Thus, we are by no means restricted to the name of Jehoram here. The temptation is to opt for the name of a familiar figure, such as Jehoram, son of Ahab. However, there are a number of issues that should be borne in mind.
\nFirst, the damaged context of the fragments does not allow us to pinpoint a name with any accuracy. Since the name following the lexeme \u05d1\u05e8 has not been preserved, the immediate context is unknown. Schniedewind\u2019s report that the slightest remains of an \u2019aleph are preserved at the very edge of Line B7 are not sustainable after vigorous observation of the fragment, and so they must be discounted. Thus, we must not be restricted to searching for a candidate whose father\u2019s name began with the letter \u2019aleph.
\nThis means that any reconstruction of the name here in Line B7 will inevitably be affected by external factors, such as how one interprets the history of Syria-Palestine in Iron Age II. This is perhaps the most critical issue of all and unfortunately infuses the interpretation with a large degree of reader subjectivity. In order to maintain some objectivity, we must balance an interpretation with other factors. The palaeographic dating and archaeological context of the inscription should provide the chronological bracket in which we place the inscription, and hence our selection of personages available from that time. However, we must not be fooled into necessarily selecting a figure known to us from other sources, for we may have a reference here to a previously unknown figure. We should certainly test the names of figures known to us, but they should not restrict us. Furthermore, although it is most likely that a personage mentioned with their patronymic was a king, it is not necessarily the case. Again, the damaged context hinders us from a full knowledge of the circumstances at this point in the Tel Dan Inscription.
\nWith this platform, we may proceed to unravel the mystery of the name in Line B7 as best we can. Palaeographically and archaeologically, the inscription was assigned to c. 800 BCE. Biran and Naveh\u2019s suggestion that the name here in Line B7 is that of \u2018Jehoram, son of Ahab\u2019 is unlikely in this case, for Jehoram is supposed to have lived approximately half a century earlier, being assassinated by Jehu in 841 BCE. The only possible way of reading \u2018Jehoram, son of Ahab\u2019 here is to understand it as a very distant retrospection or as part of a multiple patronymic. The difficulty with this last suggestion is that Jehoram was the last member of the Omride Dynasty, which was allegedly annihilated by Jehu. Thus, it is virtually impossible that a descendant of Jehoram survived to be mentioned in the Tel Dan Inscription. With the annihilation of the Omride Dynasty, it is difficult to ascertain what relevance Jehoram, son of Ahab, had to the author who lived some time later.
\nThe only resolution to this problem is that the author is Hazael, who died in c. 798 BCE. His reign, which began in c. 843 BCE, provides an overlap with that of Jehoram of barely two years. Yet, Hazael is characterized in biblical literature as a usurper, confirmed by his description in Assyrian records as \u2018the son of a nobody\u2019. Na\u2019aman has recently hypothesized that Hazael was actually a royal prince of Beth-Re\u1e25ob, the son of Ba\u2019asa, who is known from Shalmaneser III\u2019s account of the Battle of Qarqar (c. 853 BCE). As such, claims Na\u2019aman, Hazael would have had no hesitation in referring to his father, Ba\u2019asa. Thus, Na\u2019aman sees Hazael as the author of the Tel Dan Inscription.
\nHowever, this theory is based on a number of questionable interpretations of key texts. First, seeing Hazael as the son of Ba\u2019asa does not account for Hazael being labelled in Assyrian records as \u2018the son of a nobody\u2019. Second, Na\u2019aman\u2019s claim that the biblical account of \u2018Bar Hadad\u2019s\u2019 death (2 Kgs 8:15) does not necessarily implicate Hazael as his murderer is unsustainable. Third, reinterpreting the text of Hazael\u2019s Horse Ornament from Samos, as Na\u2019aman does, to identify Hazael\u2019s origins in \u2018Amqi\u2019 (i.e. the Beqa\u2018Valley) does not automatically elevate him to the status of a hereditary royal prince.
\nThat Na\u2019aman\u2019s theory is based on a number of highly questionable connections and reinterpretations is clearly seen. Therefore, Hazael should still be considered a usurper who had no hereditary claims to royalty. This disqualifies him from being the author of the Tel Dan Inscription. Any connection, then, with Jehoram, son of Ahab, is lost, meaning we cannot identify the figure mentioned in Line B7 as Jehoram of Israel.
\nAnother alternative is to consider \u05e8\u05dd\u2014as part of the name of the king of Bayt-Dawid. If we identify the rulers of Bayt-Dawid with the rulers of Jerusalem, then Jehoram, son of Jehoshaphat comes to mind. However, once again, this Jehoram lived approximately half a century before the production of the Tel Dan Inscription. A multiple patronymic, though, would resolve this problem, for the same dynasty ruled Jerusalem (and Judah) for centuries. The leading candidate in that case would be Jehoash, grandson of Jehoram, who reigned in Jerusalem from c. 835\u2013796 BCE. Thus, we might have a remnant of the name [\u05d9\u05d4\u05d5\u05d0\u05e9\u00b7\u05d1\u05e8\u00b7\u05d0\u05d7\u05d6\u05d9\u05d4\u05d5\u00b7\u05d1\u05e8 \u05d9\u05d4\u05d5]\u05e8\u05dd\u00b7\u05d1\u05e8\u00b7[\u05d9\u05d4\u05d5\u05e9\u05e4\u05d8] (\u2018[Jehoash, son of Ahaziah, son of Jeho]ram, son of [Jehoshaphat\u2019]).
\nYet we must ask what purpose a multiple patronymic served the author here. Kings are known to cite their own multiple patronymics, 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, it is unlikely that his multiple patronymic was cited by the author. Therefore, we must have doubts about the validity of restoring the multiple patronymic of Jehoash, son of Ahaziah, here, especially as it would require the inclusion of at least three predecessors\u2019 names.
\nThere 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 \u05e8\u05dd\u2014, 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.
\nThe name \u05d0\u05dc\u05e8\u05dd (\u2018\u2019Ilram\u2019) is known to have been a popular name in Ammon, perhaps even the name of a seventh-century king. The element \u05e8\u05dd\u2014is also to be seen in the name of the Edomite king, Aiarammu, known from Sennacherib\u2019s annals (c. 701 BCE), and the popular Tyrian name, A\u1e25iram or \u1e24iram. Noll mentions this latter possibility as a plausible option for reconstructing this name in Line B7. 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\u2019s suggestion simply underlines the fact that the element \u05e8\u05dd\u2014was common in names all over the Levant, thus leaving us no closer to finding a solution.
\nBecking mentions the possibility that the name in Line B7 may be that of \u2019Adonleram, steward of the royal house in Hamath. \u2019Adonleram\u2019s name is known to us from the inscribed bricks from Hamath. The title given to him is \u05e1\u05db\u05df[\u00b7\u05d1]\u05d9\u05ea\u00b7\u05de\u05dc\u05db\u05d4. Yet, although \u05e1\u05db\u05df is predominantly used to denote a governor, the fact that \u2019Adonleram is called \u2018governor of the royal house\u2019 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 \u05d0\u05e9\u05e8\u00b7\u05e2\u05dc\u00b7\u05d4\u05d1\u05d9\u05ea (\u2018who is over the house\u2019), 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 connection between \u2019Adonleram and a figure named in Line B7 is possible, but unlikely.
\nPerhaps 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 \u05dc\u05de\u05db\u05d1\u05e8\u05dd (\u2018belonging to Makbirram\u2019). It is dated stratigraphically to the early eighth century, making it synchronous with the Tel Dan Inscription. Although 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.
\nHowever, a fair degree of uncertainty must be maintained because of the fragmentary nature of the context and the relative ambiguity surrounding Makbirram\u2019s person. Nevertheless, it must be said that Makbirram presents as good a match as any for the name in Line B7 in terms of the epigraphical, archaeological and historical data. Thus, the name in Line B7 may tentatively be identified with Makbirram of Hazor. It should, however, 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.<\/p>\n
Line B8
\n\u05d9\u05af\u05d4\u05d5\u00b7\u05d1\u05e8. 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 \u05d4\u05d5 do not make a meaningful lexeme on their own. Official Aramaic transcribes the 3.masc.sg. independent personal pronoun as \u05d4\u05d5, but this is not the case in Old Aramaic, where the pronoun is written \u05d4\u05d0. Hence it is not the case here in the Tel Dan Inscription.
\nThe only choice we have is to join the first three extant letters to form the element \u05d9\u05d4\u05d5. The fact that this is followed by the lexeme \u05d1\u05e8 (\u2018son\u2019) means we must view \u05d9\u05d4\u05d5 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 \u05d9\u05d4\u05d5\u05d4. The apocopated form, \u05d9\u05d4\u05d5, appears as the theophoric element of numerous names originating in Israel and Judah where Yahwistic cults prevailed. The implication for Line B8 is that the figure mentioned here is almost definitely an Israelite or a Judaean.
\nFor the century between 852 BCE and 753 BCE, no king of Israel possessed a name with the final element being the theophoric element, \u05d9\u05d4\u05d5. 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 [\u05d0\u05de\u05e6]\u05d9\u05af\u05d4\u05d5\u00b7\u05d1\u05e8[\u00b7\u05d9\u05d5\u05d0\u05e9] (\u2018Amaziah ben-Joash\u2019). Once again, though, this is only a probability and not a certainty. 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 reconstruction of his name here is the most likely option, seeing that royal personages are likely to be named in such monumental lapidary inscriptions.<\/p>\n
Language of the Inscription<\/p>\n
From the text of the three fragments, it is clear that the language of the Tel Dan Inscription is Old Aramaic. This conclusion is based on the following combination of factors:<\/p>\n
1. The representation of the original phoneme \u1e0f by the consonant qoph (\u05e7), demonstrated in the noun \u05d0\u05e8\u05e7 (\u2018land\/ground\u2019) in Lines A4 and A10.
\n2. The retention of taw in the two conjugations of the verb \u05e7\u05ea\u05dc in Lines A6 and A8.
\n3. The indication of the masculine plural absolute by \u05df\u2014, seen in the initial word of Line B6, [\u2026]\u05e2\u05af\u05df\u05af.
\n4. The 1.com.sg. independent personal pronoun, \u05d0\u05e0\u05d4, in Line A5.
\n5. The indication of 3.masc.sg. pronominal suffix by \u05d4\u2014, seen in Line B2 with the lexeme [\u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4 in Line B2.
\n6. The use of the preposition \u05e7\u05d3\u05dd, seen with a suffix in Line A5.
\n7. The nun of the preposition \u05de\u05df not assimilating to the following consonant due to the entire preposition not being prefixed to the following word in Line B5.
\n8. The definite object marker, \u05d0\u05d9\u05ea, in Lines A9\u201310.
\n9. The prothesis of \u2019aleph in the noun \u05d0\u05e9\u05af\u05dd\u05af in Line A9.
\n10. The postpositive nun in the masculine singular adjective \u05d0\u05d7\u05e8\u05df in Line A11.
\n11. The use of characteristically Aramaic vocabulary:
\na. The noun \u05d1\u05e8 (\u2018son of\u2019. Lines B7, B8);
\nb. The verbal root \u221a\u05d4\u05d5\u05da (\u2018to go\u2019, Lines A3, A5);
\nc. The verbal root \u221a\u05e1\u05dc\u05e7 (\u2018to go up\u2019. Line A2).
\n12. The elision of lamed in the Pe\u2018al imperfect conjugation of \u221a\u05e1\u05dc\u05e7, demonstrated in Line A2 with the lexeme \u05d9\u05e1\u05e7\u05af.<\/p>\n
The combination of all these factors makes it certain that the language we are dealing with is Old Aramaic.
\nA large gap in the data is the lack of any definite article. This, however, should not be taken as a characteristic of the inscription\u2019s language, but as a result of the inscription\u2019s breakage into fragments. That is, no definite article is attested on the three fragments available to us, but there is nothing 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, \u05d0\u2014, would have been used as a normal feature of the language, Old Aramaic, or, if not expressly 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\u2019s Horse Ornament and of the Sam\u2019alian 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\u2019s dialect to that of the Sefir\u00e9 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.<\/p>\n
Grammatical Survey<\/p>\n
Orthography<\/p>\n
Graphemes. A total of 21 different graphemes are attested on all three fragments as a whole. The only grapheme that remains unattested is \u1e6deth. The single occurrence of \u1e63adhe is also quite damaged, though there is no doubt as to its identification.<\/p>\n
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:<\/p>\n
1. The grapheme qoph (\u05e7) represents the phonemes q and \u1e0f.
\na. The representation of phoneme q is demonstrated by the words \u05e7\u05d3\u05dd (Line A4), \u05e7\u05d3\u05de\u05d9 (Line A5), \u05d5\u05d0\u05e7\u05ea\u05dc (Line A6), \u05d5\u05e7\u05af\u05ea\u05af\u05dc\u05af[\u05ea\u05d4] (Line A8) and [\u05de]\u05d0\u05e4\u05e7 (Line B5).
\nb. The representation of phoneme \u1e0f is demonstrated by the words \u05d1\u05d0\u05e8\u05e7 (Line A4) and \u05d0\u05e8\u05e7 (Line A10).
\n2. The grapheme \u0161in (\u05e9) represents the phonemes \u015b and \u0161.
\na. The representation of phoneme \u015b is demonstrated by the words \u05ea\u05af\u05e9\u05e8 (Line A1) and \u05d9\u05e9[\u05e8\u05d0\u05dc] (Lines A8, A12).
\nb. The representation of phoneme \u0161 is demonstrated by the words \u05d5\u05d9\u05e9\u05e8\u05d1 (Line A3), \u05e4\u05e8\u05e9 (Line A7), \u05d5\u05d0\u05e9\u05af\u05dd\u05af (Line A9) and \u05e9\u05af\u05d1\u05e2\u05af (Line B5). The grapheme \u0161in (\u05e9) may also have served to represent the phoneme \u1e6f, but such an instance is unattested among the three extant fragments.<\/p>\n
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:<\/p>\n
1. At the end of a word, the grapheme heh can represent either the vowel \u00e2 or the vowel \u00ea in an open syllable.
\na. The representation of vowel \u00e2 in a final open syllable is seen in the word \u05d0\u05e0\u05d4 (Line A5).
\nb. The representation of vowel \u00ea in a final open syllable is seen in the word [\u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4 (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 pronominal suffix did actually represent the vowel \u00ea. However, no such words are attested among the small portion of text preserved on our three fragments.
\n2. At the end of a word, the grapheme waw represents a long \u2018u\u2019-type vowel in an open syllable.
\na. The representation of long vowel \u00fb in a final open syllable is seen in the word [\u05d0\u05de\u05e6]\u05d9\u05af\u05d4\u05d5 (Line B8). We might expect that the final vowel of the afformative for 3.masc.pl. verbs was also represented by the grapheme waw and indicative of long vowel \u00fb. However, no such verbs are attested among the three extant fragments.
\nb. The representation of long vowel \u00f4 is unattested among the three extant fragments.
\n3. At the end of a word, the grapheme yodh can represent either the vowel \u00ea (or \u00e2) or \u00ee in an open syllable.
\na. The representation of vowel \u00ea in a final open syllable is seen in the words [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af (Line A2), \u05d5\u05d0\u05dc\u05e4\u05d9 (Line A7) and \u05d0\u05af\u05e1\u05e8\u05d9 (Line B6). We might expect that this vowel was properly represented by the grapheme heh (see 1.a. above). When the vowel \u00ea 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 vowels in the final syllables of construct nouns, we must consider them diphthongs.
\nb. The representation of vowel \u00ee in a final open syllable is seen in the words \u05d0\u05d1\u05d9 (Lines A2, A3), \u05de\u05dc\u05db\u05d9 (Line A6) and \u05de\u05dc\u05db\u05d9\u05af (Line B3). In these instances, the final yodh is indicative of the 1.com.sg. pronominal suffix.
\n4. Although the grapheme \u2019aleph is not attested at the end of a word in an open syllable, we might expect that it could have represented the vowel a, either in the emphatic ending of a definite article (assumed to be post-positional \u05d0\u2013), or in particular words such as the particle of negation, \u05dc\u05d0. However, no instances of such words are attested among the extant fragments.<\/p>\n
Diphthongs. The orthography of the extant fragments demonstrates that diphthongs were preserved and had not contracted, either in an open or a closed syllable:<\/p>\n
1. The diphthong ay is represented by the grapheme yodh.
\na. The diphthong ay is seen in an open syllable in the word \u05e7\u05d3\u05de\u05d9 (Line A5). In this particular instance, the diphthong represents the 1.com.sg. suffixed pronoun.
\nb. The diphthong ay is seen in a closed syllable in the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Line A9).
\nc. If the diphthong ay had not yet contracted to \u00ea in the final syllable of construct nouns, then the grapheme yodh also indicated the masc.pl. construct.
\n2. The diphthong aw is unattested among the extant fragments, but must have been represented by the grapheme waw in other parts of the original inscription. It had been previously thought that diphthong aw was preserved in the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Line A9), but this must be revised in light of the preceding interpretation of the meaning and phonology of this word.<\/p>\n
Matres Lectionis. The orthography of the text indicates that matres lectionis 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 discussion of diphthongs above). To this end, the masc.pl. ending was indicated by \u05df\u2014(see [\u2026]\u05e2\u05af\u05df\u05af, Line B6), not by \u05d9\u05df\u2014. On analogy with the Zakkur texts, we may surmise that plene spelling may have been used on occasion. However, from the sample of text available to us, this cannot be confirmed and would actually appear unlikely.<\/p>\n
Pronouns. The extant portions of the inscription attest the following pronouns.<\/p>\n
Independent personal pronouns. Only two independent personal pronouns are attested among the extant fragments:<\/p>\n
a. The 1.com.sg. form, \u05d0\u05e0\u05d4 (Line A5).
\nb. The 3.masc.pl. form, \u05d4\u05dd (Line A10). This particular pronoun is a special case since it is treated syntactically as a pronominal suffix. Orthographically, however, it stands independently of the preceding noun.<\/p>\n
Suffixed personal pronouns. Only two pronominal suffixes are attested among the extant fragments:<\/p>\n
1. 3.masc.sg. (\u05d4-):
\na. Attested only after verbal forms: [\u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4 (Line B2).
\n2. 1.com.sg. (\u05d9-):
\na. After nominal forms: \u05d0\u05d1\u05d9 (Lines A2, A3), \u05de\u05dc\u05db\u05d9 (Line A6), \u05de\u05dc\u05db\u05d9\u05af (Line B3).
\nb. After prepositions: \u05e7\u05d3\u05de\u05d9 (Line A5).
\nc. After object marker: \u05d0\u05af[\u05d9\u05ea\u05d9] (Line B4).<\/p>\n
Other pronouns. 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.<\/p>\n
Nouns. The inscription contains numerous nouns, many of which are undamaged. Some, however, require reconstruction, while others are identifiable as nouns but defy specific reconstruction.<\/p>\n
Forms. The following types of nominal forms are attested among the three extant fragments:<\/p>\n
a. Type ql: \u05d0\u05d1 (Lines A2, A3), \u05d1\u05e8 (Lines B7, B8).
\nb. Type qtl: \u05d0\u05dc\u05e3 (Lines A6 [reconstructed], A7), \u05d0\u05e8\u05e7 (Lines A4, A10), \u05d0\u05e9\u05af\u05dd\u05af (Line A9), \u05de\u05dc\u05da (Lines A6, A8\u20139, B3), \u05de\u05e6\u05af\u05e8 (Line A13), \u05e4\u05e8\u05e9 (Line A7), [\u05e8]\u05db\u05d1 (Lines A6\u20137), \u05e9\u05af\u05d1\u05e2\u05af (Line B5), [\u2026]\u05e2\u05af\u05df\u05af (damaged pl. noun. Line B6).<\/p>\n
Inflexion of the noun. The limited data of the nominal forms attested in the extant fragments does not allow us to fully develop a paradigm for the inflexion of nouns. All the nominal forms attested are masculine, with the sole exception of the noun \u05d0\u05e8\u05e7. However, as this noun was considered feminine by nature rather than by inflexion, the data is incomplete. For some of the nouns, particularly those representing numerals, we may surmise feminine forms by appealing to other Old Aramaic texts for parallels. However, most of the nouns attested among the extant fragments are masculine by nature rather than inflexion. For these nouns, no feminine equivalent can be derived. Another gap in the data is that no nominal forms are attested in the emphatic state with the definite article.
\nThus, our paradigm is very sparse with attested forms. For those forms that are not attested among the three extant fragments of the Tel Dan Inscription, a paradigmatic noun \u05e7\u05ea\u05dc is employed and the form is underlined and preceded by an asterisk (*). The inflexion of these forms is based on the conventions of Old Aramaic, but it must be realized that we have no definitive evidence of their inflexion among the extant fragments of the Tel Dan Inscription. As the paradigm for feminine forms would be entirely theoretical, there being no evidence whatsoever for their inflexion among the extant fragments of the Tel Dan Inscription, the feminine paradigm is not included.<\/p>\n
Table 6.1. \u2018Paradigm\u2019 of masculine nouns, based on the information from the Tel Dan Inscription.<\/p>\n
Masculine Forms
\nSingular
\nPlural
\nAbsolute
\n\u05e4\u05e8\u05e9 (Line A7)
\n[\u2026]\u05e2\u05af\u05df\u05af (Line B6) *\u05e7\u05ea\u05dc\u05df
\nConstruct
\n\u05de\u05dc\u05da (Line A8)
\n\u05d0\u05dc\u05e4\u05d9\u05af (Line A7)
\nEmphatic
\n*\u05e7\u05ea\u05dc\u05d0
\n*\u05e7\u05ea\u05dc\u05d9\u05d0<\/p>\n
Adjectives. There is only one attestation of an adjectival form among the extant fragments. This is the lexeme \u05d0\u05d7\u05e8\u05df (Line A11). This is a masculine singular form, but the noun that it modifies is no longer extant. Reasonably, however, we may assume congruence between the noun and the adjective pertaining to number and gender. The inflexion of the adjective, therefore, would have followed the inflexion of the modified noun in these aspects.<\/p>\n
Prepositions. Six prepositions are attested among the extant fragments. These can be put into two classifications:<\/p>\n
1. Independent Prepositions:
\na. \u05d0\u05dc (Line A3).
\nb. \u05de\u05df\u05af (Line B5).
\nc. \u05e2\u05dc (Lines A1 [reconstructed], A12, A13).
\nd. \u05e7\u05d3\u05dd (Line A5).<\/p>\n
Of these independent prepositions, only \u05e7\u05d3\u05dd is attested with a pronominal suffix (\u05e7\u05d3\u05de\u05d9).<\/p>\n
2. Prefixed Prepositions:
\na. \u2014\u05d1 (Lines A2 [reconstructed], A3 [reconstructed], A4, B2 [twice; once reconstructed].
\nb. \u2014\u05dc (Line A10, A11).<\/p>\n
In addition to these, we may add the particle marking the direct object, \u05d0\u05d9\u05ea. This is placed immediately before the object of a clause or can have a suffixed pronoun to indicate the object.<\/p>\n
Conjunctions. Only one conjunction is attested in the Tel Dan fragments\u2014namely, the copula,\u2014\u05d5. This is prefixed to both verbal and nominal forms. When prefixed to verbal forms, it does not serve the function of a waw consecutive.<\/p>\n
Verbs. There are five verbal stems attested on the Tel Dan fragments:<\/p>\n
1. Pe\u2018al.
\n2. Pe\u2018il.
\n3. Pa\u2018el.
\n4. Haph\u2018el.
\n5. Ithpe\u2018el.<\/p>\n
Within these, we may recognize the following categories of aspect:<\/p>\n
1. Perfect.
\n2. Imperfect.
\n3. Infinitive.
\n4. Jussive.
\n5. Participle.<\/p>\n
These are the conjugations and aspects attested in the fragments. There is no evidence of, nor any need to see, the presence of consecutive imperfects derived from a preterital form.
\nThe Ithpe\u2018el is attested here only in the infinitive within a damaged context (Line B2). It is, therefore, difficult to ascertain whether the preformative of the stem in the perfect and\/or infinitive conjugations was formed with a\u2014\u05d4\u05ea particle or an\u2014\u05d4\u05ea particle. An Ithpa\u2018al perfect form is attested in Bar Rakib A.14 with a\u2014\u05d4\u05ea particle (\u05d4\u05ea\u05e0\u05d0\u05d1\u05d5), suggesting that a\u2014\u05d4\u05ea particle is plausible for an Ithpe\u2018el form. All other attested forms of Ithpe\u2018el or Ithpa\u2018al are in the imperfect conjugation, some being ambiguous due to damaged contexts. I have opted for a\u2014\u05d4\u05ea particle on the basis of the attestation of the Ithpa\u2018al perfect form in Bar Rakib.<\/p>\n
Adverbs. No adverbial forms are attested among the extant fragments of the Tel Dan Inscription.<\/p>\n
Glossary of the Inscription<\/p>\n
List of Words
\n\u05d0\u05d1 \u2018father\u2019, n.masc.\u2014sg. + 1.com.sg. sfx, \u05d0\u05d1\u05d9, Lines A2, A3.
\n\u05d0\u05d7\u05e8\u05df \u2018other, another\u2019, adj.\u2014masc.sg., Line A11.
\n\u05d0\u05d9\u05ea nota acc.\u2014[\u05d0]\u05d9\u05ea, Lines A9\u201310; \u05d0\u05d9\u05ea (+ 1.com.sg. sfx), \u05d0\u05af[\u05d9\u05ea\u05d9], Line B4.
\n\u05d0\u05dc \u2018to\u2019, prep.\u2014Line A3.
\n\u05d0\u05dc\u05e3 \u2018thousand\u2019, n.masc.\u2014pl. cstr. +\u2014\u05d5, \u05d5\u05d0\u05dc\u05e4\u05d9, Line A7; { pl. cstr., [\u05d0\u05dc\u05e4\u05d9]}, Line A6.
\n\u05d0\u05e0\u05d4 \u2018I\u2019, pron. 1.com.sg.\u2014Line A5.
\n\u05d0\u05e1\u05e8 \u2018to bind, capture\u2019, vb.\u2014Pe\u2018il, participle masc.pl. cstr. {or Pe\u2018al, participle masc.pl. cstr.}, \u05d0\u05af\u05e1\u05e8\u05d9, Line B6.
\n\u05d0\u05e4\u05e7 \u2018to be brave, determined\u2019, vb.\u2014Pa\u2018el, participle masc.sg., [\u05de]\u05d0\u05e4\u05e7, Line B5.
\n\u05d0\u05e8\u05d0\u05dc \u2018(altar) hearth\u2019, n. masc.\u2014sg. abs., [\u05d0]\u05e8\u05d0\u05dc, Lines A3\u20134.
\n\u05d0\u05e8\u05e7 (1) \u2018land\u2019, n.fem.\u2014sg. cstr., Line A10.\u2551 (2) \u2018ground\u2019\u2014sg. cstr. +\u2014\u05d1, \u05d1\u05d0\u05e8\u05e7, Line A4.
\n\u05d0\u05e9 \u2018man\u2019, n.masc. (coll.)\u2014sg. abs., \u05d0[\u05e9], Line B6.
\n\u05d0\u05e9\u05dd \u2018name\u2019, n.masc.\u2014sg. cstr {or sg. abs.} +\u2014\u05d5, \u05d5\u05d0\u05e9\u05af\u05dd\u05af, Line A9.
\n\u2014\u05d1 \u2018in, at, on\u2019, pfx.prep.\u2014\u05d1 + n., Lines (A2), {A3}, A4; [\u05d1] + inf. + 3.masc.sg. sfx., [\u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4, Line B2.
\n\u05d1\u05e8 \u2018son\u2019, n.masc.\u2014sg. cstr., \u05d1\u05e8, Lines B7, B8.
\n\u05d2\u05d6\u05e8 \u2018to cut\u2019, vb.\u2014Pe\u2018al, masc. (person and number unknown) +\u2014\u05d5, \u05d5\u05d2\u05d6\u05e8\u05af[\u2026], Line B1.
\n\u05d4\u05d5\u05da \u2018to go, walk\u2019, vb.\u2014Pe\u2018al, impft 3.masc.sg. +\u2014\u05d5, \u05d5\u05d9\u05d4\u05da, Line A5; juss. 3.masc.sg., \u05d9\u05d4\u05da, Line A3.
\n\u05d4\u05dd \u2018they, them\u2019 pron. 3.masc.sg.\u2014Line A10.
\n\u2014\u05d5 \u2018and\u2019, pfx.conj.\u2014\u05d5 + pft., Line A8; \u05d5 + impft., Lines A3, A5, A6; \u05d5 +\u2014\u05dc + inf. (?), Line A11; \u05d5 + n., Lines (A2), A7, A9; \u05d5 + {?}, Line A6.
\n\u05d7\u05e1\u05d3 \u2018piety, pious act\u2019, n.masc.\u2014sg. +\u2014\u05d5 +\u2014\u05d1 + 1.com.sg. sfx., [\u05d5\u05d1\u05d7]\u05e1\u05af[\u05d3]\u05d9\u05af
\n\u05d9\u05d5\u05dd \u2018day\u2019, n.masc.\u2014pl. cstr., [\u05d9\u05d5\u05de]\u05d9, Lines A5\u20136.
\n\u05db\u05dc \u2018all\u2019, n.masc.\u2014{sg. cstr. +\u2014\u05d1, [\u05d1\u05db\u05dc]}, Line A3.
\n\u2014\u05dc \u2018to\u2019, pfx.prep.\u2014\u05dc + inf.(?), Line A11; \u05dc + {?}, Line A10.
\n\u05dc\u05d7\u05dd \u2018to fight, battle\u2019, vb.\u2014Ithpe\u2018el, inf (+\u2014\u05d1) + 3.masc.sg. sfx., [\u05d1\u05d4]\u05ea\u05af\u05dc\u05d7\u05de\u05d4, Line B2.
\n\u05de\u05dc\u05da \u2018to reign\u2019, vb.\u2014Pe\u2018al, pft 3.masc.sg., [\u05de]\u05dc\u05da, Lines A11\u201312. Haph\u2018el, pft 3.masc.sg., \u05d4\u05de\u05dc\u05da, Line B4.
\n\u05de\u05dc\u05da \u2018King\u2019, n.masc.\u2014sg. abs., \u05de\u05dc[\u05da], Line A6; sg. cstr., \u05de\u05dc\u05da, Line A8; [\u05de\u05dc]\u05da, Lines A8\u20139; sg. + 1.com.sg. sfx, \u05de\u05dc\u05db\u05d9\u05af, Line B3; pl. abs., [\u05de\u05dc\u05db\u05df], Line [B5].
\n\u05de\u05dc\u05da \u2018reign\u2019, n.masc.\u2014sg. + 1.com.sg. sfx., \u05de\u05dc\u05db\u05d9\u05af, Line A6.
\n\u05de\u05df \u2018from\u2019, prep.\u2014\u05de\u05df\u05af, Line B5.
\n\u05de\u05e6\u05e8 \u2018siege\u2019, n. masc.\u2014sg. abs., \u05de\u05e6\u05af\u05e8, Line A13.
\n\u05e1\u05dc\u05e7 \u2018to go up, ascend\u2019, vb.\u2014Pe\u2018al, juss. 3.masc.sg., \u05d9\u05e1\u05e7\u05af, Line A2.
\n\u05e2\u05dc (1) \u2018against\u2019, prep.\u2014\u05e2\u05dc, Line A13. \u2551 (2) \u2018over\u2019\u2014\u05e2\u05dc, Line A12; \u05e2[\u05dc], Line A1.
\n\u05e2\u05dc\u05dc \u2018to come\u2019, vb.\u2014Pe\u2018al, impft 3.masc.sg. +\u2014\u05d5, \u05d5\u05d9\u05e2\u05dc, Line B3.
\n\u05e4\u05e8\u05e9 \u2018cavalry, horsemen\u2019, n.masc.\u2014sg. abs., \u05e4\u05e8\u05e9, Line A7.
\n\u05e7\u05d3\u05dd \u2018ancient\u2019, adj.\u2014masc.sg., Line A4.
\n\u05e7\u05d3\u05dd \u2018before\u2019, prep.\u2014\u05e7\u05d3\u05dd + 1.com.sg. sfx., \u05e7\u05de\u05d3\u05d9, Line A5.
\n\u05e7\u05ea\u05dc \u2018to slay\u2019, vb.\u2014Pe\u2018al, pft 1.com.sg. +\u2014\u05d5 {+ 3.masc.sg. sfx.}, \u05d5\u05e7\u05af\u05ea\u05af\u05dc\u05af[\u05ea\u05d4], Line A8; impft 1.com.sg., +\u2014\u05d5, \u05d5\u05d0\u05e7\u05ea\u05dc, Line A6.
\n\u05e8\u05db\u05d1 \u2018chariotry, chariots\u2019, n.masc.\u2014sg. abs., [\u05e8]\u05db\u05d1, Lines A6\u20137.
\n\u05e9\u05d1\u05e2 \u2018seven\u2019, n.\u2014masc.sg. abs., \u05e9\u05af\u05d1\u05e2\u05af, Line B5.
\n\u05e9\u05d9\u05dd \u2018to put, lay\u2019, vb.\u2014Pe\u2018al, pft 1.com.sg., [\u05e9\u05de\u05ea], Line A12.
\n\u05e9\u05db\u05d1 \u2018to repose, lie\u2019, vb.\u2014Pe\u2018al, impft 3.masc.sg. {or juss. 3.masc.sg.} +\u2014\u05d5, \u05d5\u05d9\u05e9\u05db\u05d1, Line A3.
\n\u05e9\u05e8\u05e8 \u2018to rule\u2019, vb.\u2014Pe\u2018al, impft 2.masc.sg., \u05ea\u05af\u05e9\u05e8, Line A1.<\/p>\n
Names of Persons and Deities
\n\u05d0\u05dc\u05d1\u05d9\u05ea\u05d0\u05dc \u2018El-Baytel\u2019 [i.e. \u2018The Bethel-stone, El\u2019, a developed form of the deity El], n.pr.dei.masc.\u2014\u05d0\u05dc\u05af\u05d1\u05d9[\u05ea\u05d0\u05dc], Line A4.
\n\u05d0\u05de\u05e6\u05d9\u05d4\u05d5 \u2018Amaziah\u2019, n.pr.pers.masc.\u2014[\u05d0\u05de\u05e6]\u05d9\u05af\u05d4\u05d5, Line B8.
\n\u05d4\u05d3\u05d3 \u2018Hadad\u2019, n.pr.dei.masc.\u2014Lines A5, [B3], B4.
\n\u05d9\u05d4\u05d5\u05d0\u05d7\u05d6 \u2018Jehoahaz\u2019, n.pr.pers.masc.\u2014Lines [A11], [B8].
\n\u05d9(\u05d4)\u05d5\u05d0\u05e9 \u2018Jehoash, Joash\u2019, n.pr.pers.masc.\u2014Lines [A11], [B8].<\/p>\n
Names of Places
\n\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 \u2018Bayt-Dawid\u2019, n.pr.loc.a [name of the city state centred on Jerasalem]\u2014Line A9.
\n\u05d9\u05e9\u05e8\u05d0\u05dc \u2018Israel\u2019, n.pr.gent.\u2014Line A8; \u05d9\u05e9[\u05e8\u05d0\u05dc], Line A12.
\n\u05e9\u05de\u05e8\u05df \u2018Samaria\u2019, n.pr.loc.\u2014[\u05e9\u05de\u05e8\u05df], Line A13.<\/p>\n
Chapter 7<\/p>\n
HISTORICAL COMMENTARY<\/p>\n
Introductory Remarks<\/p>\n
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.
\nWith the Tel Dan Inscription we have a primary written source. Furthermore, 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 information 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 secondary sources and employ them with the appropriate contingent weight.<\/p>\n
Historical Considerations<\/p>\n
The Author of the Inscription
\nAlthough 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\u2019s city gates, our author lived sometime c. 800 BCE. During this time, the city of Dan was under Aramaean influence. 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, located on the fringe of the Aramaean region, just 70 km from Damascus, would most definitely have come under the sway of the potent Aramaean state as Hazael extended his suzerainty.
\nThe archaeological and epigraphical evidence from Dan bears this out. During the reign of Hazael in the second half of the ninth century BCE, Dan underwent a major building phase: the second of the three building phases in Stratum III. The previous building phase was contemporary with the Omride Dynasty, but it is difficult to attribute the construction to the Omrides. Distinctly Israelite remains do not appear until the third construction phase, dated to the eighth century BCE. For example, Proto-Aeolic capitals only appear in the layer of debris that covered the large piazza built at the gate complex. Since the piazza was built in the second construction phase (second half of the ninth century BCE), the Proto-Aeolic capital is necessarily later. Dated to the same time as the capital is the stamped handle of an amphora unearthed in a room (No. 9024) to the immediate west of the shrine complex in Tel Dan\u2019s northwest corner (Area T). The handle is stamped \u05dc\u05e2\u05de\u05d3\u05d9\u05d5 (\u2018belonging to \u2018Immadiyo\u2019)\u2014a distinctly Hebrew name of Israelite origin. Prior to this, there is a distinct Aramaic-speaking culture in Dan, as attested by the Aramaic Tel Dan Inscription from the second phase of construction, along with a sherd from a bowl inscribed with the Aramaic word \u05dc\u05d8\u05d1[\u05d7]\u05d9\u05d0 (\u2018belonging to the butch[er]s\u2019). This Aramaic-speaking influence can even be seen as far south as \u2018Ein Gev on the eastern shore of the Sea of Galilee, where ajar inscribed with the word \u05dc\u05e9\u05e7\u05d9\u05d0 (\u2018belonging to the cupbearers\u2019) was unearthed during excavations. In light of the conquest of Dan prior to the first construction phase (presumably at the hands of Bar Hadad I) and the continuity between the first and second construction phases, it seems probable that an Aramaic-speaking culture prevailed in Dan throughout the majority of the ninth century BCE, being replaced by an Israelite culture only in the early eighth century BCE.
\nThe emerging picture, therefore, is one in which Dan was already in the orbit of Damascus before Hazael even ascended the throne in c. 843 BCE. The renaissance of an Israelite culture at Dan at the beginning of the eighth century BCE must be attributed to the successful exploits of Jehoash and his son, Jeroboam II. Damascene hegemony over Dan is further enforced by the polished language of the Tel Dan Inscription (Old Aramaic).
\nWesselius\u2019s suggestion that the author of the Tel Dan Inscription is to be identified with Jehu of Israel flounders here on numerous grounds. Wesselius\u2019s hypothesis fails to consider the archaeological landscape of Tel Dan and its telling implications for reconstructing the political and cultural climate of both Israel, Damascus, and more specifically, of Dan itself during the ninth and eighth centuries BCE. Rather, Wesselius\u2019s theory is based solely on reconciling the text of the Tel Dan Inscription with the biblical narratives of Jehu. Not only are there a number of dubious assertions in this hypothesis, but Wesselius also fails to address how and why Jehu erected a stele for himself at Dan so far from the Israelite heartland, especially if (as Wesselius insists) Jehu was a vassal of Hazael\u2019s ever-growing Damascene state.
\nMoreover, Wesselius\u2019s hypothesis rests on circumstantial evidence and too many uncertain possibilities. Wesselius proceeds largely on the basis that since the inscription mentions the slaying of two kings, then it is worth investigating whether the king boasting of the slayings is Jehu who, in the biblical narrative, is behind the slaying of two kings. This, of course, is a reasonable suggestion. However, that the Tel Dan Inscription mentions the killing of two kings is conjecture. Although it may originally have done so, the most we can say with certainty is that the extant fragments mention the slaying of one king. The second king may well have been connected to this in some way, but the Tel Dan fragments do not spell this out. Furthermore, Wesselius insists that it is only if the Jehu hypothesis is overturned that we can then proceed with other hypotheses. This is unnecessarily restrictive, particularly if much of the evidence happens to be circumstantial. Indeed, much of the evidence to date has been circumstantial due to the incorrect arrangement of the inscription\u2019s fragments. Since Wesselius\u2019s hypothesis relies on this incorrect arrangement, the basis of the evidence for the Jehu hypothesis no longer stands. Coupled with the fact that the hypothesis does not square well with the archaeological evidence, it is highly unlikely, if not impossible, that Jehu authored the Tel Dan Inscription.
\nLines A11\u201312 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.
\nMost 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 identified the two names in Lines B7 and B8 as those of \u2018[Jeho]ram son of [Ahab]\u2019 and \u2018[Ahaz]iah son of[Jehoram]\u2019, 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.
\nData 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 influence over at least the majority of states in Syria-Palestine. Hazael\u2019s death occurred in c. 799 BCE\u2014the very time to which the archaeological and palaeographical evidence point for the date of the inscription\u2019s 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.
\nThe first few lines of Fragment A, however, prevent us from identifying Hazael as the author of the Tel Dan Inscription. Hazael was most certainly a usurper of the throne of Aram-Damascus. This fact is clearly conveyed in other corroborated sources. The Assyrian annals call Hazael \u2018the son of a nobody\u2019, while the biblical text portrays him as the murderer of Bar Hadad II (Hadad-\u2018I\u1e0fri), his predecessor (2 Kgs 8:15). It is unlikely, indeed impossible, that Hazael would have referred to his own father with such terms as are used in the Tel Dan Inscription, for evidently Hazael\u2019s father was not the previous king of Aram-Damascus.
\nThis was a point first noticed by Biran and Naveh in their publication of Fragment A. However, this crucial datum was not properly dealt with in Biran and Naveh\u2019s revised theory once Fragment B was published. Biran and Naveh speculated that Hazael might have been part of the royal dynasty he usurped. Yet, this speculation does not negate the fact that Hazael\u2019s father was not the king whose throne he usurped. For Biran and Naveh, who proposed that Hazael authored the Tel Dan Inscription, this presents a major flaw in their theory. Indeed, this flaw serves to demonstrate that their theory has been largely forced onto the evidence. That is, Biran and Naveh\u2019s theory attempts to mould the evidence into a shape that it simply cannot hold. This fact alone seriously undermines their entire historical reconstruction for the inscription.
\nRecognizing this key difficulty in identifying Hazael as the author, Schniedewind and Zuckerman propose an ingenious solution. They hypothesize that Hazael was the leader of a political group aligned with worship of the deity El, whereas his predecessor on the throne, Hadad-\u2018I\u1e0fri, was part of a faction owing allegiance to Hadad. This sets up an intriguing picture of the religio-political situation of Damascus. However, there are three fundamental flaws in this theory. First, in Line A5 the author indicates allegiance towards Hadad. According to Schniedewind and Zuckerman\u2019s theory, this must either see Hazael as a turncoat or a blatant liar. Neither, of course, can be totally dismissed. However, the more critical flaws lie in the fact that Schniedewind and Zuckerman follow the old arrangement of the fragments and the readings dependent on them, and a failure to examine the archaeological strata of Tel Dan.
\nThe \u2018complete confidence\u2019 shown by Margalit in identifying Hazael as the author of the Tel Dan Inscription is, therefore, totally unfounded. Margalit\u2019s conclusion is based entirely on the (erroneous) identification of the two names in Lines B7 and B8 with Jehoram of Israel and Ahaziah of Judah. As has been noted in previous chapters, this identification is practically impossible on archaeological and epigraphical grounds. Margalit deals with the difficulty of Hazael\u2019s apparent reference to his \u2018father\u2019 in the following manner:<\/p>\n
Hazael, who was probably not in direct line of succession, and may even have been a commoner, credits Hadad\u2019s favour for his accession to the throne as successor of Hadad-Izri whom he (disingenuously?) calls \u2018father\u2019 in the inscription.<\/p>\n
This treatment of the issue is quite inadequate. Furthermore, the tone and punctuation which Margalit uses shows that not even he is actually completely confident in explaining the difficulty of positing Hazael as the author of the Tel Dan Inscription. Based on the text proposed by Biran and Naveh, Margalit claims that the 1.com.sg. independent pronoun in the (awkward) reconstruction [\u05d5]\u05d9\u05af\u05d4\u05de\u05dc\u05da\u00b7\u05d4\u05d3\u05d3[\u00b7]\u05d0[\u05ea\u05d4\u05ea\u00b7]\u05d0\u05e0\u05d4 emphasizes that it was Hazael, and not any other pretender, whom Hadad put on the throne of Aram-Damascus. However, since the pronoun appears on Fragment A while the rest of the expression appears on Fragment B, Margalit\u2019s theory falls apart. It requires the fragments to be arranged in the old configuration, which has been demonstrated as no longer sustainable.
\nUnfortunately, Margalit goes on to make other conclusions about the content of the inscription based on the false premises that Hazael was the author and that the names in Lines B7 and B8 are those of Jehoram of Israel and Ahaziah of Judah. Not only do these conclusions have no real foundation, but the methodology employed to reach them is decidedly backward. If these conclusions are reached by means of reverse approaches to research, then there must be a distinct lack of confidence in identifying Hazael as the author of the Tel Dan Inscription. We must be led by the time-frame of the inscription indicated by the archaeological data, and the fact that Hazael was certainly a usurper.
\nSasson, who appears to have followed both Margalit\u2019s and Yamada\u2019s theories, deals with the issue in a similar way:<\/p>\n
Hazael could have been of royal blood but not necessarily in line of succession 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\u2019s father.<\/p>\n
Sasson\u2019s comments, a repeat of Margalit\u2019s previous suggestions, appears to be an attempt simply to conform the notion that Hazael was actually the author of the Tel Dan Inscription with the evidence from the Assyrian and biblical sources. Sasson\u2019s suggestion that Hazael was not actually a commoner but perhaps the son of a minor wife or concubine fails to realize the implication of the Assyrian label, mar la mammana (\u2018son of a nobody\u2019). For all the propagandistic purposes of Shalmaneser III\u2019s texts, the implication in calling Hazael the \u2018son of a nobody\u2019 is that Hazael was most certainly not of royal stock. If this was merely a piece of propaganda designed to discredit Hazael without any shred of truth in it, then we must ask why Shalmaneser did not employ the same \u2018name-calling\u2019 against other adversaries, especially against those of the various coalitions that stood up to him far more successfully than Hazael ever did. The reason must undoubtedly be that those other adversaries were kings of legitimate royal stock whereas Hazael was not. Hazael may have held a position in the court of his predecessor, but he was not his predecessor\u2019s son.
\nSasson also follows Na\u2019aman\u2019s suggestion that there was no foul play in the death of Hazael\u2019s predecessor. Dealing with the biblical narrative on Hazael\u2019s accession to the throne, Sasson hypothesizes that either Hazael was nursing his predecessor when he died, or Hazael\u2019s predecessor killed himself either accidentally or deliberately. Throughout these two suggestions, Sasson maintains that Hazael was a very close confidante of his predecessor, sharing \u2018a relationship almost, if not actually, as close as that of a son to his father\u2019. In actual fact, the biblical narrative merely portrays Hazael as an emissary. To suggest anything beyond that is to venture even beyond the source that Sasson tries to extract from. Yet, even if we allow for the understanding that no foul play was involved in the death of Hazael\u2019s predecessor, we must ask why the direct heir did not then assume the throne of Aram-Damascus. By Sasson\u2019s own admission, Hazael was evidently not in the direct line of succession. Unless the death of Hazael\u2019s predecessor was accompanied simultaneously by the death of his direct heir(s)\u2014either his sons and\/or those in a more direct line of succession\u2014we must understand Hazael\u2019s appropriation of the Damascene throne as the result of deliberate court intrigue. In other words, Hazael was a usurper in the fullest sense of the word.
\nLemaire follows a more coherent path in his treatment of the issue, though it too falls short of accounting for the difficulties inherent in seeing Hazael as the author of the Tel Dan Inscription. Lemaire appeals to Middle Kingdom Egypt and the Neo-Assyrian Empire to make a case that it was not unusual for a usurper to refer to his predecessor as \u2018father\u2019. He also appeals to the text of 1 Sam. 24:11 in which David refers to Saul as \u2018my father\u2019. This latter example, though, may be dismissed for David is clearly not portrayed as the \u2018usurper\u2019 of Saul\u2019s throne. In fact, David is portrayed as Saul\u2019s son-in-law who eventually fills the vacuum left by Saul\u2019s son and successor, Ish-bosheth. The same may well have been the case with the example of Tiglath-Pileser III who could have inherited the kingship of Assyria from a power vacuum left in the wake of a revolt in Nimrud. This leaves Lemaire\u2019s example of Middle Kingdom Egypt. However, the relevance of Pharaonic ideologies and the machinations of Egyptian politics in the Middle Bronze Age to the situation of Hazael in Iron Age II is dubious. Nonetheless, for all that the examples of Egypt and Assyria may entail, Lemaire\u2019s argument is based on the assumption that Hazael was the author of the Tel Dan Inscription. Not only does this not comport with the archaeological and epigraphical data of the Tel Dan Inscription, it also represents a backward methodology. Lemaire seeks not to enquire as to whom the author of the Tel Dan Inscription might have been, but to find evidence supporting the assertion that Hazael was the author. The laudable terms used by the author of the Tel Dan Inscription in reference to his father\u2014and he speaks of himself with equal praise\u2014do not permit the interpretation that Hazael was the author of the stele. Such an interpretation disregards the weight of numerous independent sources.
\nIn further support of his hypothesis, Lemaire cites Line B4 and the factitive use of \u221a\u05de\u05dc\u05da. This, claims Lemaire, indicates that the \u2018succession between Hadadezer [Hadad-\u2018i\u1e0fri] and Hazael was not natural\u2019. Once again, Lemaire\u2019s comments are based on the assumption that Hazael was the author of the Tel Dan Inscription. Disregarding this, however, his claim that the factitive use of \u221a\u05de\u05dc\u05da in Line B4 indicates an unusual or problematic succession at first appears valid. However, we must note the position of this phrase in the entire inscription, as well as the implications of its syntax.
\nFirst, Lemaire\u2019s transcription of the relevant phrase is based on the text reconstructed by Biran and Naveh, and so is incorrect. The relevant verb is not a consecutive imperfect ([\u05d5]\u05d9\u05de\u05dc\u05da), as Lemaire proposes, but a simple perfect (\u05d4\u05de\u05dc\u05da). Second, contrary to Lemaire\u2019s arrangement of the fragments, it has been shown that Fragment B belongs in the lower portions of the original stele. If this phrase was referring to an unnatural succession that had to be legitimated through divine appeal, then it is quite strange that it appears so far down in the inscription. One expects that legitimacy for the throne would be stated at the outset. Although Lemaire\u2019s understanding of the arrangement puts this phrase at the beginning of the stele, the arrangement is unsustainable. Third, the relevant phrase (\u05d4\u05de\u05dc\u05da\u00b7\u05d4\u05d3\u05d3\u00b7\u05d0\u05af[\u05d9\u05ea\u05d9]) follows a reference to the author\u2019s king, presumably Hadad, in Line B3. Lines B4\u20135 are, in turn, bracketed by references to war and captives. Thus, it is with reference to war that we must understand the author\u2019s claim that Hadad made him king. That is, the author appears to have claimed that his special task as king was specifically to make war and that it was for this divinely ordained task that Hadad made him king\u2014a fairly normative Levantine ideology. Line B4, then, appears not to be a statement legitimating a dynasty, but a declaration for holy war. Since, then, the author did succeed his father to the throne, the author cannot have been Hazael. It must be said that if it requires backward methodology based on a defunct reading of the inscription to validate the claim that Hazael was the author, then we must abandon the notion.
\nOnly one other king suffices for identification as the author of the Tel Dan Inscription, and that is Bar Hadad, Hazael\u2019s son and successor. This Bar Hadad has been enumerated as the third, or even fourth, monarch of that name to have worn the crown of Aram-Damascus. However, for reasons that will be discussed below, Bar Hadad, the son of Hazael, should be enumerated as only the second monarch of Aram-Damascus with that name. Thus, I will refer to him as Bar Hadad II.
\nIt is not precisely known when Bar Hadad II succeeded his father as king of Aram-Damascus, though we can estimate an approximate date. The Tell el-Rimah Stele mentions that Adad-Nirari III received a large tribute from a certain Mari of the land of Damascus. This information is corroborated by the Nimrud Slab. It would appear that this occurred in the year 796 BCE, which in the Eponym Chronicle is described as the year \u2018against Man\u1e63uate\u2019. Man\u1e63uate appears to have been an Assyrian designation for the Beqa\u2018 Valley in the Lebanon. The year 796 BCE is the only time during which Adad-Nirari III could have campaigned against Damascus and received tribute from its king, one of whom was Jehoash of Israel who had not come to the throne until c. 798 BCE. Adad-Nirari III is likely to have been the \u2018saviour\u2019 who relieved Israel in 2 Kgs 13:5 from the oppression of Hazael and his son, Bar Hadad II. This also means we should identify Mari from the Tell el-Rimah Stele with Bar Hadad II. Thus, the indications are that Bar Hadad II had been on the throne of Aram-Damascus for at least a few years before 796 BCE.
\nBar Hadad II\u2019s death occurred some time before 773 BCE, for in that year, Shamshi-ilu, field marshal under Shalmaneser IV, campaigned against Hadianu of Damascus and received tribute from him. This campaign is mentioned in the Pazarcik stele, while in the Eponym Chronicle, the year is referred to as the year \u2018against Damascus\u2019. The only other clue for the date of Bar Hadad II\u2019s death comes from the Zakkur Stele. Although no concrete dates can be gleaned from the stele, the political situation it describes allows us to give an approximate date. In its description of the siege against Ha\u1e0frakh, the stele places Bar Hadad II at the head of 16 armies and their kings. According to Ahlstr\u00f6m, the siege could not have occurred after 796 BCE when Adad-Nirari III exacted tribute from Bar Hadad II. However, Assyria\u2019s preoccupation with its northern enemy, Urartu, in the latter years of Adad-Nirari III and the early years of Shalmaneser IV, could also have occasioned Bar Hadad\u2019s alliance and the siege of Ha\u1e0frakh after some years of respite from Assyria\u2019s opposition. The Zakkur Stele was written at least some years after the siege of Ha\u1e0frakh since Zakkur mentions his vast building projects undertaken after the struggle with Bar Hadad II and his alliance. Since Hadianu had come to the throne of Aram-Damascus by 773 BCE, we should date the siege of Ha\u1e0frakh sometime between c. 784\u2013775 BCE\u2014that is, the last years of Adad-Nirari III and the first years of Shalmaneser IV. Thus, Bar Hadad II was still on the throne of Aram-Damascus at this time.
\nThe pottery assemblage gathered from the third construction phase at the southern Iron Age gate of Tel Dan was dated to the first half of the eighth century BCE. Since Hadianu ruled Aram-Damascus towards the end of this period (c. 775\u2013750 BCE), he could not have been the author of the Tel Dan Inscription, which was recycled as building matter for the third phase of construction at Dan\u2019s city gate. This third construction phase must have occurred in the first few years of the eighth century BCE. Therefore, Bar Hadad II is the only real candidate for the author of the Tel Dan Inscription.<\/p>\n
The Author\u2019s Father
\nThe breaking of the Tel Dan Inscription must logically have occurred immediately prior to the third phase of construction at Dan\u2019s city gate. The hammer blow that destroyed the inscription is still evident as the semi-circular recess along the left edge of Fragment A. Thus, it was a very deliberate act of vandalism that destroyed the original stele. This destruction of the stele should be taken as an act of reprisal aimed against Bar Hadad II himself. A corollary of this is that the stele had been erected as an act of triumph or glorification of Bar Hadad II. The excellent condition of the inscription\u2019s incisions and the date of the script demand that there was not a great period of time between these two acts.
\nIt is not surprising that Bar Hadad II praises the exploits and memory of his father, Hazael, in the first few lines of the inscription\u2019s extant portions. 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\u2019s death. Both the Hazael Ivory and the Horse Ornament from Samos both refer to Hazael as \u05de\u05e8\u05d0\u05df (\u2018our lord\u2019). 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 Stele\u2014Mari.
\nAt the time of Hazael\u2019s death, Aram-Damascus had taken control of the region\u2019s 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\u2019s 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\u2019s capture of Transjordanian territories, giving him the King\u2019s Highway and subsequently the rich trade from Arabia. Amos\u2019s oracles condemn Damascus for a violent conquest of Gilead (Amos 1:3\u20135), and Israel\u2019s boast of having captured Lo-Debar and Qarn\u00e1yim (Amos 6:13) implies that these Transjordanian 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\u201333, 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.
\nThe biblical record also has Hazael campaigning against Gath and seizing it (2 Kgs 12:17\u201318). 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\u2019s conquest at the hands of Hazael (late ninth century BCE) or the hands of Sargon II (712 BCE). If it refers to the latter, then we must date the reference in Amos to a period well after Amos\u2019s ministry (c. 760 BCE). Since Gath\u2019s 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\u2019s downfall as legendary derives from that time. Although Gath may simply have been known for its many downfalls, further support for this late date is the fact that the downfall of Kalneh and Hamath are also described in Amos 6:2 as comparable to that of Gath. Both these cities fell to Tiglath-Pileser III in 738 BCE. It is possible that Amos 6:2 refers to the conquests of Kalneh, Hamath and Gath by Hazael, but this is much less likely.
\nNevertheless, it is not inconceivable that Hazael did indeed campaign on the Philistine Plain. Israel was certainly weak enough politically, economically and militarily for this to have occurred. The tradition of a harsh oppression of Israel by Damascus, particularly at the hands of Hazael, is quite pervasive 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. Whether this destruction can be attributed 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.
\nThis 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 \u2018Way of the Sea\u2019) between Egypt and Palestine. 2 Kings 12:17\u201318 also has Hazael initiating a campaign towards Jerusalem. If this indeed occurred, it was presumably to affect control over the road along the watershed of Palestine\u2019s 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 Jerusalem, it implies that he was able to gain effective control of these routes through the submission of Joash of Judah. If Hazael had actually campaigned that far south, his control of these regions must have been nominal or indirect, purely because of the seasonal economy of Damascus (as opposed to the perennial economies of the Mesopotamian and Egyptian superpowers) and the distance of these southern localities from his seat of rule. Only if Hazael employed regional governors could his control of such southern territories have been strong. In any case, it seems clear from the evidence that Hazael\u2019s ascendancy and economy were fuelled by his secure possession of the King\u2019s Highway in Transjordan and the routes branching to the north of Damascus.
\nThe throne that Bar Hadad II inherited from his father was a prestigious and powerful one. Hazael\u2019s military, political and economic success would undoubtedly have been seen as a blessing from the gods, and Hazael himself would have been considered as their minion. The wealth captured by Hazael would have ensured the prosperity and ascendancy of Damascus during his lifetime. What the Tel Dan Inscription shows is Bar Hadad II\u2019s desire to consecrate and perpetuate the memory of his deceased father through a kind of ancestor cult enacted at various shrines belonging to the deity El-Baytel.
\nLine A1 of the Tel Dan Inscription implies that Bar Hadad II was Hazael\u2019s named successor. In Lines A4\u20135, Bar Hadad starts to recount incidents from his own reign. In these lines, he seems to liken himself to his father whom he earlier describes as a pious man (Line A2). Bar Hadad obviously saw himself as a favourite of the gods, in particular, Hadad, to whom he ascribes the successes of his own military campaigns with the phrase, \u2018so Hadad would go before me\u2019. It is reasonable to suggest that Bar Hadad lived in the shadow of his father\u2019s success and desired to emulate his achievements. He attained at least one success worthy of mention.<\/p>\n
The King of Israel
\nOne of the successes that Bar Hadad experienced is mentioned first in generalized terms and then furnished with specifics. In Lines A6\u20137, Bar Hadad makes a name for himself as a slayer of kings and a defeater of armies. In Line A8, he gives us the specific detail that he killed the king of Israel. Three kings of Israel were contemporaries of Bar Hadad during his reign\u2014namely, Jehoahaz, Jehoash and Jeroboam II, all successive rulers from the Jehu dynasty. The biblical records, however, do not recall that any of these three Israelite kings were killed by Bar Hadad. Jeroboam II died in 753 BCE\u2014too late to be the king killed by Bar Hadad since Bar Hadad died c. 775 BCE. Jeroboam\u2019s father, Jehoash, died in 782 BCE. This also seems slightly too late for Jehoash to have been the king killed by Bar Hadad because the pottery assemblage from the third construction phase at Tel Dan predates this slightly and the Tel Dan Inscription was recycled during this construction phase. This leaves Jehoahaz who died in 798 BCE. This date for Jehoahaz\u2019s death fits perfectly with the archaeological and palaeographical data of the Tel Dan Inscription. Therefore, we can conclude that Jehoahaz ben-Jehu was the king of Israel whom Bar Hadad claims to have killed in Line A8.
\nIn Lines A11\u201312, 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\u2019s death (2 Kgs 13:9). This is corroborated by the Tell er-Rimah Stele, which records that Adad-Nirari III received the tribute of \u2018I\u0101su\u2019 (Joash) of Samaria. This must have occurred during 796 BCE when Adad-Nirari III campaigned against Man\u1e63uate in the Lebanese Beqa\u2018, just two years after Jehoash\u2019s accession to the Israelite throne in Samaria. Hence, I have restored the name \u2018Jehoash\u2019 or \u2018Joash\u2019 to the end of Line A11.
\nThe biblical record of Jehoahaz\u2019s 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 time\u2014that is, conflict between the states of Aram-Damascus and Israel:<\/p>\n
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\u2019s eyes and followed the offences of Jeroboam ben-Nebat, who had made Israel offend. He did not stray from them. So Yahweh\u2019s 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\u20133)<\/p>\n
This account of conflict between the monarchs of the two Levantine states squares well with Bar Hadad\u2019s 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:<\/p>\n
Yet Jehoahaz beseeched Yahweh\u2019s person and Yahweh listened to him because he saw Israel\u2019s 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\u20135)<\/p>\n
This preserves the memory of Adad-Nirari III\u2019s campaign to Man\u1e63uate in 796 BCE in which Bar Hadad (\u2018Mari\u2019 in the Tell er-Rimah Stele) was forced to pay a hefty tribute to the Assyrian king. The Saba\u2019a 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. The Tell er-Rimah Stele puts the tribute payment at 2000 talents of silver, 1000 talents of copper, 2000 talents of iron and 3000 multicoloured and plain linen garments. The Nimrud Slab puts the tribute at 2300 talents of silver, 20 talents of gold, 3000 talents of copper, 5000 talents of iron, coloured woollen and linen garments, an ivory bed, a couch inlaid and embossed with ivory, and an immeasurable amount of Bar Hadad\u2019s property and goods. 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\u2019s reprieve from Aram-Damascus came in Jehoahaz\u2019s lifetime, should be altered. The Tell er-Rimah Stele and the Eponym Chronicle show that the reprieve came during the reign of his son, Jehoash.
\nBar Hadad had inherited from his father, Hazael, a large sphere of influence in Syria-Palestine. After Adad-Nirari III\u2019s 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\u2019s incorporation into Israel. The biblical record corroborates this:<\/p>\n
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\u201325)<\/p>\n
Jehoash of Israel was able to take advantage of Aram-Damascus\u2019 weakness after the campaign of Adad-Nirari III in 796 BCE. The tribute exacted on Bar Hadad rendered him weak against the emerging and determined new king of Israel, Jehoash. The Israelite king\u2019s subsequent recovery of the territories taken by Hazael (and Bar Hadad?) must have occurred quite soon after Adad-Nirari\u2019s departure from the region. One of the cities that Jehoash \u2018retrieved\u2019 must have been Dan, for the stele erected there by Bar Hadad\u2014the Tel Dan Inscription\u2014was knocked down at this time. Therefore, the fracturing of the inscription and its recycling in the third construction phase at Dan must be associated with Jehoash\u2019s victories over Bar Hadad and the economic advantages that these victories wrought for Israel. The recapture of Transjordanian territories by Israel would have given Jehoash (and his son, Jeroboam II) control of strategic stretches of both the King\u2019s Highway and the coastal road from Egypt. Bar Hadad\u2019s line of supply from these routes would have been effectively choked by his losses against Jehoash.<\/p>\n
The Toponym \u2018Bayt-Dawid\u2019
\nAs a toponym, Bayt-Dawid can be understood as either (1) the name of a city, similar to \u2018Bethel\u2019, or (2) the name of a state, similar to \u2018Beth-Re\u1e25ob\u2019. The fact that this geographical entity had a king implies some sort of state. This does not necessarily lead us any closer to identifying the location, as we could be dealing with either a state occupying a region known as Bayt-Dawid, or a city-state centred on a city called Bayt-Dawid. Our course of action should be to understand how exactly the name \u2018Bayt-Dawid\u2019 was understood by the author of the Tel Dan Inscription, Bar Hadad II. Thus, we must look at the contexts surrounding the occurrence of the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in Line A9.
\nThe apparent alliance between the king of Israel and the king of Bayt-Dawid is implied in two ways by Bar Hadad. First, they are both mentioned close together in successive lines (Lines A8 and A9). Second, Bar Hadad groups the land of these two kings together. This means that Bar Hadad either viewed the land of these two kings as a single political entity, or, more likely, as a single terrain. It is unlikely that any other kings were mentioned in addition to the king of Israel and the king of Bayt-Dawid at this point in the inscription. References to any other kings would all have to have occurred within the space of Lines A8\u20139. It is clear that the king of Israel was Bar Hadad\u2019s primary opponent since his death is recounted as a great victory for Bar Hadad who is then obligated to mention his successor. Thus, the entire alliance must be recalled between the initial reference to the king of Israel (Line A8) and the later reference to his successor (Lines A11\u201312). The reference to \u2018their land\u2019 (Line A10), however, implies that by Line A10, Bar Hadad had listed all the parties in alliance with the king of Israel and had begun to explain what he inflicted on \u2018their land\u2019. There is hardly enough space, therefore, for more kings to be mentioned here.
\nThis apparent alliance between the king of Israel and the king of Bayt-Dawid, as well as Bar Hadad\u2019s reference to \u2018their land\u2019, must mean that the region or town of Bayt-Dawid was an immediate neighbour to the kingdom of Israel. Presumably, this means the two states shared a common border and a single terrain. This information places Bayt-Dawid somewhere 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.
\nAs mentioned previously, no entity named \u2018Bayt-Dawid\u2019 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\u2019s 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 \u2018Bayt-Dawid\u2019 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.
\nBeth 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. 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 \u05dc\u05de\u05dc\u05da (\u2018belonging to the king\u2019) 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 numerous sites throughout Judah. Many of them also bear the name of a city, though only four cities are attested from these stamped jar\u2014namely, Hebron, Socoh, Ziph and the enigmatic \u05de\u05de\u05e9\u05ea. It is quite easy to read these stamps as construct expressions, such as \u2018Belonging to the king of Hebron\u2019 or \u2018Belonging to the king of Socoh\u2019.
\nHowever, numerous factors count against this interpretation. First, these storage jars are all of the same essential type and the letters \u05dc\u05de\u05dc\u05da 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. The iconography, palaeography, epigraphy and archaeological context of these jars point to a date towards the end of the eighth century BCE. 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. Also, the clay that the jars themselves are made of appears to have a provenance in or around Jerusalem and the region around Mareshah and Moreshet-Gath in the Shephelah. The enigmatic name \u05de\u05de\u05e9\u05ea has variously been interpreted as a reference to \u2018Jerusalem\u2019 or the government of Judah centred in Jerusalem. To this we may add that it might be a reference to Mareshah or Moreshet-Gath (perhaps 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 \u05de\u05de\u05e9\u05ea 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. Thus, at the end of the eighth century BCE, Beth Shemesh was not a city-state, but part of the regional state of Judah.
\nYet, 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\u201314). 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 experienced 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.
\nThis 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 \u2018residency\u2019, but even by their own admission, there is nothing about the structure to suggest it was a palace or even a governor\u2019s residence. Furthermore, the wall of Beth Shemesh was severely breached during the destruction of the Stratum IV city, probably at the hands of the Philistines. These breaches were repaired with rubble during subsequent levels of occupation and were made considerably narrower (1 m in thickness) than the original wall (2.4 m in thickness), or were simply built over in the construction of houses. This could hardly be called advanced or formidable fortification. Thus, if there had been a city-state at Beth Shemesh at this time it was a very weak one. In any case, the abundance of Philistine ware from Stratum III of Tell er-Rumeilah, which we date from the late twelfth to the late tenth centuries BCE, indicates that Beth Shemesh was well and truly in the orbit of Philistia at this time. The less prosperous city of Stratum II and its various phases which are of relevance to the time of the Tel Dan Inscription\u2019s manufacture, highlights the unlikelihood of any state administration centred at Beth Shemesh.
\nHowever, the weak defences and the lack of palatial structures at Beth Shemesh must be offset by the presence of industrial installations for the processing of olives and wine, a dyeing industry and an apparent increase in population in Iron II. These indicate that Beth Shemesh was a thriving town, though not as prosperous as the previous Philistine-administered city, and certainly not an independent city-state.
\nThe statement in 2 Kgs 14:11 that Beth Shemesh belonged to Judah further supports the notion that Beth Shemesh was part of a larger regional state. This territorial claim in the biblical text should alert us to the pro-Jerusalem and pro-Judaean bias of the author-compilers. That Amaziah of Judah is described as drawing Jehoash of Israel into battle at Beth Shemesh implies that Beth Shemesh was in Israel\u2019s possession and that Judah had to wrest the town from Israel in order to press its claim. That the recollection of this incident corroborates quite well with the archaeological and geopolitical landscape of the early eighth century BCE must verify at least some degree of accuracy in the biblical account. All the sources, therefore, point to Beth Shemesh being part of a larger state entity centred on another city at the time in question. It is most unlikely, therefore, that Beth Shemesh was the Bayt-Dawid of the Tel Dan Inscription.
\nJerusalem, however, fits the evidence remarkably well. The biblical text gives the constant impression that Jerusalem was always the capital of a larger state called \u2018Judah\u2019. This was certainly the case in the late eighth century BCE and the late seventh to early sixth century BCE. However, Jerusalem did not attain the status of a metropolis until the mid-eighth century BCE. By the time of Sennacherib\u2019s campaign against Hezekiah (701 BCE), Jerusalem had been strongly fortified. With the city expanding into the \u2018second Quarter\u2019 (\u05d4\u05de\u05e9\u05e0\u05d4) on the Western Hill, a new \u2018broad wall\u2019 was erected to protect these new outer environs. At this time, we can see the influence of Jerusalem extending to all corners of Judah\u2019s greater area. Among the evidence for this is Sennacherib\u2019s own annals recounting his campaign in 701 BCE. Sennacherib mentions having besieged 46 of Hezekiah\u2019s cities and taking them from Hezekiah\u2019s country and giving them to various Philistine kings. This must be a reference to a regional state of Judah. Sennacherib also calls Hezekiah \u2018the Judaean\u2019 (Akkadian la\u00fadai), obviously in reference to the country that Hezekiah ruled.
\nPrior to this time, however, Jerusalem was a smaller town, confined to the ridge popularly known as the \u2018City of David\u2019 and the area roughly equating to the current Temple Mount. Amos\u2019s oracles against the nations (Amos 1:3\u20132:16) 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. Thus, we are safe in determining that in c. 760 BCE the state of Judah extended from the northern environs of Jerusalem 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\u2019s influence extended all the way to Edom, even as far as the Gulf of Aqabah. This is indeed a plausible situation considering that the former power of the region, Aram-Damascus, had been considerably weakened and its territories along the King\u2019s Highway taken from its control. Judah could easily have taken advantage of this new situation to pursue its interests in the south at this point. Effective control of the corridor between the Dead Sea and the Gulf of Aqabah only required possession of the junction between the King\u2019s Highway and the Way to the Arabah near Bozrah. Thus, with a single victory, this may have been achieved for Judah. Axiomatic to this, though, is that Judah had some type of centralized government and military organization at its disposal. The eighth century BCE economic boom in Judah implies as much. This centralized structure must have come from Jerusalem.
\nHowever, just how far Judah\u2019s influence reached into the hills west of the Central Ridge is difficult to determine. The Beth Shemesh campaign may have been an attempt on the part of Amaziah to enlarge Judah\u2019s territory, perhaps to regain traditional or ideological boundaries. Zimhoni\u2019s examination of the ceramic evidence from Lachish shows that in the second quarter of the eighth century BCE an industrial revolution began in Judah. This increase in economic activity, corresponding to Lachish III, led to the consolidation of the Judaean administration in Jerusalem, right up until Sennacherib\u2019s campaign in 701 BCE. The previous period, corresponding to Lachish IV (eighth century BCE), was characterized by a city wall and a palace-fort. This undoubtedly suggests that Lachish was an important administrative centre at this time. Yet, does it represent an autonomous government at Lachish, or is it evidence of a centralized regional government incorporating other districts of Judah?
\nAt the end of the eighth century BCE Lachish was under Jerusalem\u2019s sovereignty. 2 Kings 14:19 preserves an account of Amaziah\u2019s assassination at Lachish in c. 767 BCE. The fact that the conspirators could both murder Amaziah at Lachish and then bring his body back to Jerusalem suggests that the kings of Jerusalem did have some influence at Lachish. The beginning of Lachish III is dated to around the same time as Amaziah\u2019s assassination. The destruction of the previous city, Lachish IV, should not be attributed to a hostile takeover of Lachish by Amaziah, and thereby suggest an independent city-state at Lachish prior to that time. Rather, the destruction of Lachish IV is best explained as the result of an earthquake, presumably the one recalled in Amos 1:1 and Zech. 14:5. This quake is dated to the reign of Amaziah\u2019s son, Uzziah (Azariah). Thus, there is actually a smooth transition from the poorer economy of Lachish IV to the more affluent economy attested in Lachish III with its renewed building activity and ceramic innovations. The connection between Jerusalem and Lachish hinted at in the account of Amaziah\u2019s assassination probably indicates that Lachish had come under Jerusalem\u2019s influence by Amaziah\u2019s reign at the latest.
\nThese data suggest that there was a traditional tie between Jerusalem and the Shephelah, which the kings of Jerusalem were probably unable to enforce so long as Aram-Damascus was the dominant power of the region. With the demise of Aram-Damascus, though, Jerusalem appears to have been released from economic impediments, thus being able to actively enforce its suzerainty over Lachish. The same scenario at Beth Shemesh may well have prevailed. Indeed, it seems to have been the cause for Amaziah\u2019s conflict with Jehoash of Israel who had probably already staked his own claim on this industrially rich city in the fertile Soreq Valley. Jehoash\u2019s claim, then, stood in direct opposition to the traditional links of Beth Shemesh with Judah and hence the kings of Jerusalem.
\nAt the time the Tel Dan Inscription was written, sometime early in the reign of Jehoash of Israel, the state of Judah was probably on the verge of recovering its traditional or ideological boundaries. Jerusalem, in all likelihood, had only token sovereignty over the regions of Judah at this time. The harsh political and economic reality was that Judah had been effectively fractured into virtually autonomous districts which did not have the military or political clout to assert any real claim over other districts, especially if Aram-Damascus exercised control over the region. Instead, the economic fortunes of these districts, which could have led to military and political strength, would have been funnelled into the coffers of the Damascene kings. The relative insignificance of Judah during this time is demonstrated by Adad-Nirari III in the Nimrud Slab. In this text, the Assyrian king fails to mention Judah in the list of \u2018Amurru in its totality\u2019, though he mentions Israel (B\u00eet \u1e2aumri), Edom and Philistia. This may be a sign that, 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\u2019s Highway. If Hazael did exact tribute from Joash of Judah, then Judah must have been put into a similarly weak state.
\nIt 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 Sennacherib against Hezekiah in c. 701 BCE, Judah was again essentially reduced to these limits\u2014a fact reflected in Esarhaddon\u2019s annals, where Hezekiah\u2019s successor, Manasseh, is called \u0161arURU la\u2019udi (\u2018king of the city of Judah\u2019). 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.
\nIn 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. The Western Hill to the west of the Central (Tyropoean) Valley was not incorporated into the city until well into the reign of Hezekiah. That there was previous settlement on the Western Hill is demonstrated by the fact that Hezekiah\u2019s wall enclosing this \u2018Second Quarter\u2019 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, pointing to settlement of the Western Hill sometime during that century rather than during the previous ninth century BCE.
\nThus, the city of Jerusalem in c. 800 BCE occupied only the spur known as the \u2018City of David\u2019 or \u2018Ophel\u2019. The biblical text knows of this term, in \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019), 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\u2019s expansion under Hezekiah, the biblical text places the burial of every Judaean king in the \u2018City of David\u2019. Importantly, in 2 Sam. 5:7 it is associated with the \u2018fortress of Zion\u2019, captured by and renamed after David. It appears, therefore, that the name \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019) is associated with references to the early stages of Jerusalem\u2014namely, 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.
\nIt is my contention that the toponym \u2018Bayt-Dawid\u2019 (\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3), as used in the Tel Dan Inscription, is the equivalent of the biblical \u2018City of David\u2019 (\u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3). The noun \u05e2\u05d9\u05e8 (\u2018city\u2019) is unattested in Old Aramaic, so there is no surprise that this element has been substituted with\u2014\u05d1\u05d9\u05ea in the Old Aramaic of the Tel Dan Inscription. The reason why the noun \u05e7\u05e8\u05d9\u05d4 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. Nevertheless, we do see an interchange between the element\u2014\u05e2\u05d9\u05e8 and\u2014\u05d1\u05d9\u05ea when, in Josh. 19:41, the town of Beth Shemesh is referred to as \u05e2\u05d9\u05e8 \u05e9\u05de\u05e9 (\u2018Ir Shemesh\u2019). Thus, the interchange is certainly not unheard of.
\nSince Bayt-Dawid is the equivalent of \u2018City of David\u2019, we must therefore understand the toponym \u2018Bayt-Dawid\u2019 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, \u05d9\u05d4\u05d5\u05d3\u05d4 (\u2018Judah\u2019). Since this is not the case, \u2018Bayt-Dawid\u2019 should be considered a landed estate or a city-state that lent its name to the immediately surrounding district, rather than to the entire region of Judah. Thus, the Tel Dan Inscription confirms that Jerusalem\u2019s sovereignty over the regions of Judah was more token and ideological than an actual political reality. From the point of view of the region\u2019s most dominant power of the time, Aram-Damascus, \u2018Bayt-Dawid\u2019 was only a small feudal estate with its own lord who was in league with the larger regional state of Israel. These feudal lords of Bayt-Dawid had a close association with their feudal residence, a fortified compound on the southern spur of Jerusalem, and it was the name of this fortified compound, \u2018Bayt-Dawid\u2019 or \u2018City of David\u2019, by which the lords and their estate were known. Bayt-Dawid was, in effect, a small principality with very limited suzerainty.<\/p>\n
The King of Bayt-Dawid
\nWith the identity of Bayt-Dawid known, it becomes a relatively simple task to identify the king of Bayt-Dawid mentioned by Bar Hadad II in the Tel Dan Inscription. The king of Israel whom Bar Hadad II killed was undoubtedly Jehoahaz ben-Jehu. The king of Jerusalem at the time of Jehoahaz\u2019s death was Joash ben-Ahaziah. Since Bayt-Dawid is to be identified as Jerusalem, Joash must be the king of Bayt-Dawid referred to in the Tel Dan Inscription. Joash\u2019s reign began in c. 835 BCE after a coup to depose the Queen Dowager, Athaliah. His reign ended not long after Jehoahaz of Israel was slain by Bar Hadad; Joash was assassinated in c. 796 BCE \u2018in the House of the Millo which goes down to Silla\u2019 (2 Kgs 12:21). This House of the Millo was located in Jerusalem and may have been partly uncovered by Benjamin Mazar. Joash was succeeded by his son, Amaziah, who then began what effectively amounts to a Judaean expansionist policy in the power vacuum left in the wake of Aram-Damascus\u2019 demise. The short time span between the deaths of Jehoahaz (c. 798 BCE) and Joash (c. 796 BCE) gives us a very accurate margin with which to date the Tel Dan Inscription.<\/p>\n
The Inscription and the Biblical Text
\nThe facts we can glean from the Tel Dan Inscription confirm for us a long-held suspicion about the biblical accounts of the wars between Israel and Aram-Damascus. It has long been thought that the account of Ahab\u2019s death in battle against Bar Hadad (2 Kgs 22:29\u201337) and other narratives dealing with the Aramaean conflict have been misplaced. That is, these narratives do not refer to the reign of Ahab, but to the reigns of later Israelite monarchs. The Tel Dan Inscription confirms that these narratives actually derive from the reigns of Jehoahaz and Jehoash.
\nAhab was evidently a very strong monarch politically, economically and militarily. His ability to field the largest chariot force of the southern coalition which fought Shalmaneser III at Qarqar (853 BCE) attests to his influence, as does his connection to the Tyrian royal family through his marriage to Jezebel, daughter of Ithbaal I. At the battle of Qarqar, Ahab and his Damascene counterpart, Hadad-\u2018i\u1e0fri, both stood together in the same coalition against the Assyrian force of Shalmaneser. We may also surmise that after Ahab\u2019s death in 853 BCE, Israel continued to support the southern coalition against Shalmaneser III. Indeed, three more encounters between Assyria and the southern coalition occurred after Ahab\u2019s death: in 849 BCE, 848 BCE and 845 BCE\u2014that is, during the reign of Ahab\u2019s second son, Jehoram.
\nIt appears that a rift between Samaria and Damascus occurred only after Hazael usurped the Damascene throne in c. 843 BCE and began an expansionist policy that brought him into direct conflict with neighbouring Israel. This would be the reason why Jehoram of Israel is portrayed as warring against Hazael during which he received serious wounds (2 Kgs 8:28\u201329; 9:14\u201316). It was while Jehoram was recovering from these wounds that Jehu\u2019s coup d\u2019etat took place (in 841 BCE).
\nThe notion that Jehu\u2019s coup had been backed by Hazael (as derived from previous interpretations of the Tel Dan Inscription) does not make sense of the political situation then current. Israel still suffered terribly in the face of Hazael\u2019s expansionist policy even after Jehu usurped the Israelite throne. Hazael\u2019s reign is clearly portrayed as a time of severe oppression. Jehu\u2019s coup seems to have done nothing to resolve any differences between Samaria and Damascus. In fact, the dissolution of the anti-Assyrian coalition of the south by 841 BCE implies an obvious lack of unity and uniformity between the two major partners of the coalition, Israel and Aram-Damascus. Shalmaneser\u2019s respective treatment of Hazael and Jehu further demonstrates the fact that these two Levantine kings were clearly rivals, not allies.
\nThus, it appears that until Hazael\u2019s seizing of the Damascene throne, there was an active entente between Israel and Aram-Damascus. It is, therefore, 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. 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.<\/p>\n
1 Kings 22:1\u201340. The first episode we need to deal with is that of Ahab\u2019s 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\u2019s Highway which the Aramaeans were in possession of, but to which Israel had traditional claims. In the narrative, the king of Israel\u2019s 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\u2019s body is then conveyed to Samaria and buried there.
\nIn 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\u2019s transmission in order to discredit the Omride Dynasty, especially Ahab. The geopolitical situation at the end of Ahab\u2019s 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\u2019s death, 853 BCE, both Ahab and Hadad-\u2018i\u1e0fri of Aram-Damascus had been allies in the anti-Assyrian coalition that confronted Shalmaneser III at Qarqar and checked his advance.
\nLipi\u0144ski suggested that this narrative of Ahab\u2019s death actually recounts the death of his son, Jehoram. However, since Jehoram was assassinated in Jehu\u2019s 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. Dijkstra\u2019s article, however, appeared before the discovery of Fragments B1 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.
\nDuring Jehoahaz\u2019s lifetime, Hazael had taken possession of Israel\u2019s Transjordanian territories. Understanding 1 Kgs 22:1\u201340 in connection with Jehoahaz, it appears that upon Hazael\u2019s 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\u2019s successor, Bar Hadad II. In this battle, Jehoahaz lost his life and Israel was presumably defeated.
\nThus, the biblical account of Ahab\u2019s death has actually been derived from the memory of Jehoahaz\u2019s 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 \u2018Ahab reposed with his fathers\u2019. However, since the same formula is used to mark Jehoahaz\u2019s 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\u2019s reign and the beginning of Jehu\u2019s reign. This being the case, Ahab\u2019s 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\u2019s death at the hands of Bar Hadad II.
\nIf such a conflation of traditions did indeed occur, then it might even explain how the name of Jehoshaphat entered the narrative. The relation between the royal families of Samaria and Jerusalem through the marriage of Jehoram and Athaliah probably reflected an entente cordiale between Ahab and Jehoshaphat. It is not inconceivable, then, that Jehoshaphat accompanied Ahab to Qarqar as either a vassal or an ally. It would certainly have been politically expedient for Jehoshaphat to pursue this association with an anti-Assyrian coalition, not just to ward off Shalmaneser III, but to avoid crossing the relatively powerful Ahab and an entire coalition of 12 or more Levantine states.<\/p>\n
1 Kings 20:23\u201343. In this narrative of an Israelite victory over Aram-Damascus, the name of the Damascene king is given as \u2018Ben Hadad\u2019\u2014the Hebrew equivalent of \u2018Bar Hadad\u2019. From the Monolith Inscription of Shalmaneser III, it is known that Hadad-\u2018i\u1e0fri was the king of Aram-Damascus at this time. Thus, there is a discrepancy with the identification of the Damascene king in the biblical text. It is also telling that in this particular narrative, the king of Israel is once again anonymous. The placement of this narrative has resulted in the inference that Ahab is the king of Israel involved. Also of note is the fact that the Israelite force facing the Aramaeans was so small, it was comparable to \u2018two folds of goats\u2019 while the Aramaean army was immensely large (1 Kgs 20:27).
\nThus, the situation we have in 1 Kgs 20:23\u201343 is of an unnamed king of Israel fighting a Damascene king by the name of Bar Hadad at Apheq with the smallest of armies. The result of the battle, though, is the defeat of Bar Hadad and the subsequent return of territory to Israel, which Bar Hadad\u2019s father had captured a generation earlier. Also, the king of Israel is permitted to set up \u05d7\u05d5\u05e6\u05d5\u05ea (\u2018outposts\u2019) in Damascus. This does not fit the circumstances of Ahab at all. Apart from the discrepancy of the Damascene king\u2019s name, Ahab\u2019s army was one of the largest in the region, as evidenced by his contribution to the anti-Assyrian coalition as portrayed in the Monolith Inscription. Also, since Ahab and his Damascene counterpart were allies, we are at a loss to reasonably account for this hostility.
\nThe situation, however, does fit the circumstances of Jehoash ben-Jehoahaz remarkably well. In this instance, we see Jehoash defeating Bar Hadad II and reclaiming all the territory seized by Bar Hadad\u2019s enormously successful father, Hazael. The link with the battle site of Apheq is made when we consider the narrative of 2 Kgs 13:14\u201319. In that episode, the prophet Elisha prophesies that Jehoash would defeat Aram-Damascus three times and would have to attack the Aramaeans at Apheq. A little further on in the narrative, we are told that Jehoash did indeed defeat Bar Hadad three times and so recapture the Israelite territories which Hazael had taken from Jehoahaz (2 Kgs 13:25). We are also told in 2 Kgs 13:7 that the force of Israel during the reign of Jehoash\u2019s father, Jehoahaz, was only 50 cavalry units, ten chariotry units and ten companies (literally, \u2018thousands\u2019) of infantry soldiers. We cannot know whether these figures are accurate or not, but it definitely appears that the Israelite army in Jehoahaz\u2019s day was vastly smaller than the army in the time of Ahab. Thus, it is reasonable to suggest that the force Jehoash inherited from his father did indeed resemble \u2018two folds of goats\u2019.
\nWe see, then, how this narrative has been misplaced and taken from its original context, the first Israelite victory over Bar Hadad II under Jehoash, and put into a secondary context during Ahab\u2019s reign.<\/p>\n
1 Kings 20:1\u201322. This narrative relates a siege instigated by \u2018Ben Hadad\u2019 against Samaria during the reign of Ahab. Ben Hadad makes certain demands for hostages, which the king of Israel initially agrees to. However, once Ben Hadad increases his demands, the king of Israel refuses. An anonymous prophet predicts that the king of Israel will have victory over Ben Hadad. While Ben Hadad and his 32 accompanying kings carouse in their tents, they are attacked by a small Israelite force of seven companies (literally, \u2018thousands\u2019) and 232 city official trainees (\u05e0\u05e2\u05e8\u05d9 \u05e9\u05e8\u05d9 \u05d4\u05de\u05d3\u05d5\u05ea). Ben Hadad\u2019s siege and demands are thwarted in this successful attack, but Ben Hadad escapes on horseback with his cavalry.
\nThe 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\u201314 Ahab is named in the encounter with the anonymous prophet. It is possible that this episode has been inserted into an account of an anonymous Israelite king who was besieged in Samaria by \u2018Ben Hadad\u2019\u2014that 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 discredit 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.<\/p>\n
Catalogue of the Kings of Aram-Damascus
\nThis analysis of the biblical texts in light of the Tel Dan Inscription has highlighted that the \u2018Ben Hadad\u2019 of the Ahab narratives in Kings was not a contemporary of Ahab, but of Jehoahaz and Jehoash. It must be noted that there are two distinct Damascene kings called \u2018Ben Hadad\u2019 (i.e. Bar Hadad) in the biblical text. The first is the son of Tabrimmon, who features only in 1 Kings 15. This first \u2018Ben Hadad\u2019 is not associated with Ahab. The second \u2018Ben Hadad\u2019 is the son of Hazael, and it is with him that our interests lie.
\nBar Hadad, the son of Hazael, has commonly been enumerated as the third or fourth monarch of that name. By this reckoning, the second Bar Hadad was viewed as a contemporary of Ahab\u2014a theory based on the biblical texts that show Ahab in conflict with a certain \u2018Ben Hadad\u2019. Since the Tel Dan Inscription shows that these narratives have been conflated and derived from narratives originally involving Bar Hadad, the son of Hazael, we need to readjust our understanding of Damascene chronology and the succession of Damascene kings.
\nEssentially, our need to account for a Bar Hadad who was a contemporary of Ahab disappears in light of the Tel Dan Inscription. In other words, there was no Bar Hadad who ruled Aram-Damascus in the mid-ninth century BCE. As has long been recognized, there is a discrepancy between the text of Kings and the royal texts of Assyria. Whereas the biblical text claims that \u2018Ben Hadad\u2019 was a Damascene contemporary of Ahab, the Assyrian texts claim that the Damascene king of this time was Hadad-\u2018i\u1e0fri. This difficulty was overcome by maintaining that \u2018Ben Hadad\u2019 and Hadad-\u2018i\u1e0fri were in fact one and the same person. The Tel Dan Inscription, however, shows us that there is no longer any need to maintain this doubtful equation.
\nTwo points bear mentioning. First, in 853 BCE Hadad-\u2018i\u1e0fri i 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 year\u2014namely, 853 BCE. If there was a \u2018Ben Hadad\u2019 who was a contemporary of Ahab, then he must have reigned prior to Hadad-\u2018i\u1e0fri, 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-\u2019i\u1e0fri well before 853 BCE? The equation simply does not hold up. Therefore, the \u2018Ben Hadad\u2019 of the Ahab narratives 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.
\nWe now have a problem arising from the narrative of 2 Kgs 8:7\u201315, which twice names Hazael\u2019s predecessor as \u2018Ben Hadad\u2019. We have a number of possible solutions. First, we could understand \u2018Ben Hadad\u2019 and its Aramaic equivalent, \u2018Bar Hadad\u2019, as an honorific title for the king of Aram-Damascus, much like \u2018Pharaoh\u2019 for the king of Egypt. The difficulty with this, however, is that not every Damascene king was referred to as \u2018Bar Hadad\u2019. Hazael was certainly not referred to as \u2018Bar Hadad\u2019 and neither were any of the latter Damascene kings. In fact, epigraphic evidence clearly shows that Hazael was referred to by the title \u05de\u05e8\u05d0\u05df (\u2018Our lord\u2019). This title seems to have been used by his son, also, when we see Adad-Nirari III refer to him as \u2018Mari\u2019. Thus, all the evidence points to \u2018Bar Hadad\u2019 being a proper name as opposed to a title.
\nThe second option is to see \u2018Ben Hadad\u2019 of 2 Kgs 8:7\u201315 as Hadad-\u2018i\u1e0fri\u2019s successor who reigned for less than four years some time between 845 BCE (the last time Hadad-\u2018i\u1e0fri faced Shalmaneser III in battle) and 841 BCE (the first time Shalmaneser III encountered Hazael). Pitard has suggested this as a possibility. Attractive as this solution might be, however, it goes directly against the Assyrian epigraphic evidence. A summary inscription from the reign of Shalmaneser III clearly states that \u2018Hadad-\u2018i\u1e0fri died; Hazael, the son of a nobody, seized the throne\u2019. This statement makes the possibility that a \u2018Bar Hadad\u2019 reigned after Hadad-\u2018i\u1e0fri and before Hazael extremely slim, if not altogether impossible.
\nOur third and most plausible option is to replace the name \u2018Ben Hadad\u2019 in 2 Kgs 8:7\u201315 with that of \u2018Hadadezer\u2019 (i.e. Hadad-\u2018i\u1e0fri). This solution recognizes the confusion in the biblical text surrounding the \u2018Ben Hadad\u2019 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-\u2018i\u1e0fri. Bar Hadad, Hazael\u2019s son, should then be enumerated as the second (and last) Damascene king to bear that name.<\/p>\n
Reconstructing the Events of the Inscription
\nThe political picture that emerges from the Tel Dan Inscription is of a coalition between Jehoahaz of Israel and Joash of Bayt-Dawid against Bar Hadad II of Aram-Damascus in c. 798 BCE. Given the nature of the city-state of Bayt-Dawid and that of its northerly neighbour, Israel, it can plausibly be suggested that Israel was the senior partner in this coalition. If Hazael had indeed campaigned in the southern coastal plain at Gath (2 Kgs 12:17), both Israel and Bayt-Dawid must have been desperately weak both economically and militarily. Certainly Israel\u2019s weakness is guaranteed by their loss of Transjordanian territories to Hazael. It is not unreasonable to also suggest that Galilee had been severed from Samaria\u2019s control, especially if Hazael did indeed campaign on the southern coastal plain at Gath\u2014a situation requiring Hazael\u2019s control of the Jezreel Valley. Bayt-Dawid (i.e. Jerusalem) was still, at the time, in a dark age, even though it was a nascent state on the verge of territorial and economic expansion.
\nBar Hadad II came to the throne of Aram-Damascus after Hazael\u2019s death in c. 798 BCE. Jehoahaz appears to have used Hazael\u2019s death as a reason to attempt the recovery of Transjordanian territories. The death of Hazael marked the death of Israel\u2019s great nemesis and abuser. It would have presented a glimmer of hope for the end of Aramaean domination of Israel and Israel\u2019s perpetual weakness. Thus, it appears that Jehoahaz went into battle at Ramoth-Gilead, Joash of Bayt-Dawid by his side, with renewed vigour.
\nThe identification of Ramoth-Gilead as the location of the battle is inferred from the misplaced narrative of 1 Kgs 22:1\u201340. The choice of this town as the site of battle must undoubtedly have lain in its strategic position along the King\u2019s Highway. That Jehoahaz was able to penetrate so far east into Transjordan appears to have been a major achievement in itself. It implies that Bar Hadad, the newly crowned king of Aram-Damascus, took some time to head off Jehoahaz\u2019s advance, perhaps because he had literally just been crowned.
\nNevertheless, 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\u2019s 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).
\nThe 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\u2019s victories is implied by the fact that he mentions it immediately after praising the memory 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 Israelite 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\u201322 and 2 Kings 6\u20137 probably stems from this time. Whatever the case, Bar Hadad does boast of some raiding campaign into \u2018their land\u2019 (Line A10)\u2014that 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.
\nIf 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\u2018 at Man\u1e63uate. The biblical text also informs us that a \u2018saviour\u2019 rescued Israel from the Aramaean onslaught (2 Kgs 13:5). This \u2018saviour\u2019 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 narratives\u2014namely, 2 Kgs 7:6.
\nThe 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\u2019s 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\u2019s primary goal, and since the memory of an Aramaean siege of Samaria by \u2018Ben Hadad\u2019 is preserved in the biblical text, we can reconstruct the missing name of the city in Line A13 as \u05e9\u05de\u05e8\u05df (\u2018samaria\u2019) with a fair degree of certainty.
\nThat the siege was ultimately a failure, or at least did not lead to any wide destruction to Samaria, is evidenced by Samaria\u2019s 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 \u2018Joash of Samaria\u2019 (Akkadian: I\u0101suKURSamerin\u0101). The memory of an unsuccessful siege of Samaria by \u2018Ben Hadad\u2019 is also quite strong in the biblical text of Kings. The siege\u2019s 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\u2019s claim to have caged Hezekiah up like a bird in Jerusalem, yet without reference to Jerusalem\u2019s wall being breached or the gates penetrated. An Aramaean siege of Samaria at this time does make sense when we consider the possibility 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\u2019s willingness to check Bar Hadad, then, was the reason for Bar Hadad\u2019s retreat from Samaria, as well as the reason why Jehoash is listed as having paid tribute to Adad-Nirari III.
\nMeanwhile, Joash of Bayt-Dawid, who seems to have survived the earlier battle of Ramoth-Gilead, was assassinated in his capital city during a palace coup (2 Kgs 12:20\u201321). This would have occurred approximately two years after the battle of Ramoth-Gilead, probably in late 797 BCE or early 796 BCE\u2014that is, just before or around the same time Adad-Nirari III was campaigning in southern Syria. Joash\u2019s son, Amaziah, was installed as king of Bayt-Dawid in Joash\u2019s stead. It is Amaziah\u2019s name that appears partially preserved in Line B8 of the Tel Dan Inscription.
\nThe context of this reference appears to have been another battle. Whether this occurred before or after Bar Hadad paid a hefty tribute to Adad-Nirari III is difficult to determine. If it occurred before the Damascene\u2019s submission to Adad-Nirari, then we have a very small window of opportunity to date this battle. This is due to Adad-Nirari III\u2019s campaign at Der, east of the Tigris River, at the time of the New Year Festival of the very next year (795 BCE). Since this window of opportunity is so small, it is more probable that the battle occurred in the wake of Adad-Nirari\u2019s departure from Syria or, even more likely, that this was the battle fought between Bar Hadad and Adad-Nirari III. We note that in Line B2 of the Tel Dan Inscription, Bar Hadad refers to \u2018his fighting\u2019. Exactly whom the pronoun refers to is impossible to say because of the highly damaged context. Adad-Nirari, however, seems to be the most suitable candidate according to the time-frame.
\nExactly what role Amaziah of Bayt-Dawid played in this battle we cannot determine. Once again, this is due to the fragmentary nature of the text at this point. Exactly who the personage in Line B7 is and what role he played we also cannot tell. Although I have suggested a certain Makbirram as a possibility for this personage, we have no way of verifying this. Makbirram is known to us from excavations in Area A at Hazor. His name was found inscribed into some pottery discovered in Building 14a, a fact indicating his relative importance. Building 14a was found in Stratum VI at Hazor, but its original structure goes back to the previous level Stratum VII.
\nYadin proposed that Building 14a was a residential house and that Makbirram was a merchant who lived in it. However, there are numerous problems with this interpretation. First, that Makbirram did not actually reside in Building 14a is evidenced by the fact that another piece of pottery was found there with the name of a certain \u2018Jerob[oam] ben-Elm[atan]\u2019 written across it in ink. Yadin does not explain why the names of two seemingly unrelated persons were found on sherds in the same building. Second, Yadin proposes that part of the building consisted of shops. This certainly seems plausible. In a paved courtyard (Room 132a) within the building, Yadin found an extraordinary six millstones along with various cooking wares. Yadin states that the daily preparatory cooking must have taken place in this courtyard. However, six millstones (four bottom stones 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.
\nThis 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. 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. Interestingly, an intricate ivory cosmetic spoon was also found at the doorway. Yadin remarks that this dainty \u2018is one of the few of its kind to be found, not in palaces, but in a private house\u2019. However, we suspect that Building 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.
\nMakbirram, 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, Building 14a is located adjacent to Building 2a, which Yadin describes as \u2018the most beautifully planned and preserved building among the Israelite structures of Hazor.\u2019 Did this serve as Makbirram\u2019s residence? Whatever the case, Makbirram\u2019s 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.
\nBar 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\u2019s claim here. From the Tell er-Rimah Stele, it is clear that Bar Hadad (also known as \u2018Mari\u2019) 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 encounter 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\u2019s propaganda in the face of his waning power and a last-ditch attempt to assert his sovereignty over Galilee and Transjordan.
\nAlternatively, 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 overthrow 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\u2019s three victories over Bar Hadad attest to the fact that the Damascene king was left almost powerless 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\u2019s victories over him.
\nWhatever 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 territories. It was presumably aimed at Dan\u2019s 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) \u2018no-man\u2019s land\u2019 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\u2019s rapid demise. The subsequent revival at Dan and other key sites like Hazor demonstrates Jehoash\u2019s success at expelling Bar Hadad from both Transjordan and the Galilee. The lack of evidence for any conflagration 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\u2019s three victories at Apheq, Lo-Debar and Qarn\u00e1yim, in that these victories were enough to expel Bar Hadad and the Aramaean forces from all of Galilee and Transjordan.<\/p>\n
Date of the Inscription
\nThe Yahwistic name which is partly preserved in Line B8 is most likely that of Amaziah ben-Joash, king of Bayt-Dawid. That is, in Line B8, Bar Hadad II mentions the successor of the king of Bayt-Dawid with whom he fought in Fragment A. The succession of contemporary Israelite kings saw no monarch with a final Yahwistic theophoric element. Amaziah of Bayt-Dawid, then, is the most reasonable candidate for the person named in Line B8.
\nAmaziah came to the throne of Bayt-Dawid as the result of a palace coup against his father, Joash, which resulted in his father\u2019s assassination (2 Kgs 12:20\u201321). Since there are two years between the death of Jehoahaz of Israel in battle (c. 798 BCE) and the assassination of Joash of Bayt-Dawid (c. 796 BCE), at least two years must have transpired between the first battle mentioned in Fragment A and the second battle mentioned in Fragment B. This also means that the inscription could not have been written before c. 796 BCE.
\nThis date accords remarkably well with the archaeological context in which the fragments were found, as well as the physical condition of the fragments and other corroborated facts. The second battle, referred to in Fragment B, is either a battle between Bar Hadad II and Adad-Nirari III, or else between Bar Hadad and a coalition of southern states. Whichever is the case, 796 BCE is the earliest date at which this battle could have occurred. Therefore, 796 BCE is the earliest date at which the Tel Dan Inscription could have been produced.
\nJehoash\u2019s third victory over Bar Hadad probably brought Dan once again under Israelite control. Dan was certainly in Israelite hands during the reign of Jeroboam II (c. 782\u2013753 BCE) when Israelite influence was at its peak (cf 2 Kgs 14:25; Amos 8:14). Thus, the latest possible date for the production of the Tel Dan Inscription is the end of Jehoash\u2019s reign in c. 782 BCE. However, Jehoash probably ousted Bar Hadad from all Israelite territories not long after Bar Hadad had been humbled by Adad-Nirari III. Thus, Jehoash probably smashed the stele of the Tel Dan Inscription in the first decade of his reign. If we can, as Thiele has suggested, corroborate the beginning of Jeroboam II\u2019s co-regency in c. 792 BCE with Jehoash\u2019s battle against Amaziah, then Jehoash had probably swept to his three victories over Bar Hadad within four years, between c. 796 BCE and c. 792 BCE. Amaziah of Jerusalem\u2019s capture by Jehoash may have prompted the rise of Amaziah\u2019s son, Uzziah (i.e. \u2018Azariah\u2019) to the throne of Judah as a co-regent in his father\u2019s place. Since this occurred in c. 791 BCE, it seems reasonable to suggest that the stouche between Jehoash and Amaziah occurred close to this time. It seems reasonable, then, that by c. 792 or c. 791 BCE, Jehoash had defeated Bar Hadad three times and had smashed the stele of the Tel Dan Inscription at the end of this campaign.
\nThe demolition of the stele is clearly associated with the third construction phase at the gates of Dan. This information has come to light since excavations after the publication of all Tel Dan fragments. Dijkstra\u2019s suggestion that the original stele may have been destroyed by Hazael in an attempt to obliterate the memory of the preceding dynasty can no longer stand in light of the date demanded by the archaeological context. Dijkstra followed Biran and Naveh\u2019s lead in dating the gate complex to the ninth century BCE. However, at the time of Dijkstra\u2019s article, the three distinct construction phases at Dan\u2019s gate complex had not been discerned. Dijkstra did, however, comment that his suggestion for Hazael being behind the demolition of the original stele would be in jeopardy should the gate complex be dated to the early eighth century BCE. The recent excavations at Tel Dan have shown this to be partially the case by discerning three construction phases stretching from the mid-ninth century BCE to the early eighth century BCE.
\nJehoash\u2019s third victory over Bar Hadad and his encounter with Amaziah of Judah correspond in date perfectly with the pottery assemblage gleaned from this third construction phase\u2014namely, the very early eighth century BCE. This means that there was a maximum of five years between the production of the Tel Dan Inscription by Bar Hadad II and its subsequent destruction by Jehoash. I therefore place the production of the stele in c. 796 BCE and its destruction in c. 791 BCE.<\/p>\n
Timeline
\nThe sequence of events surrounding the production and destruction of the Tel Dan Inscription may be reconstructed as follows:<\/p>\n
c. 799 BCE:
\nHazael of Aram-Damascus dies and is succeeded by his son, Bar Hadad II. Aram-Damascus has control of the major trade routes of Syria-Palestine, rendering the states of Israel and Bayt-Dawid very weak.
\nc. 798 BCE:
\nJehoahaz ben-Jehu, king of Israel, tries to recapture the Israelite territories in Transjordan in the wake of Hazael\u2019s death. He enlists the support of Joash ben-Ahaziah, king of \u2018Bayt-Dawid\u2019 (Jerusalem), for this task. However, Bar Hadad II of Aram-Damascus defeats them at Ramoth-Gilead. In the battle, Jehoahaz is killed. He is succeeded by his son, Jehoash. To follow up his victory. Bar Hadad begins a campaign on Israelite territory in the Central Hill Country and lays siege to at least one city, probably Samaria.
\nLate 797\u2013Early 796 BCE:
\nJoash is assassinated in his capital city, \u2018Bayt-Dawid\u2019 (Jerusalem). His son, Amaziah, succeeds him. In Samaria, Jehoash appeals to Adad-Nirari III to come to his aid as Bar Hadad continues his onslaught on Israelite territory.
\nApril 796 BCE:
\nAdad-Nirari III and his Assyrian army are located at Man\u1e63uate in the Lebanese Beqa\u2018.
\nLater in 796 BCE:
\nBar Hadad lifts the siege of Samaria because of the looming Assyrian threat. Adad-Nirari III forces Bar Hadad II into submission. Bar Hadad is then required to pay a massive tribute. Jehoash of Israel also pays an undisclosed tribute to Assyria, seemingly as payment for Assyria\u2019s services in coming to Israel\u2019s aid.
\nLate 796\u2013Early 795 BCE:
\nBar Hadad erects a stele, the Tel Dan Inscription, at the gates of Dan as propaganda against his waning power. In the text of the stele, Bar Hadad portrays himself and his deceased father, Hazael, as pious men and mighty warriors favoured by the god Hadad. Bar Hadad also recounts his victory at Ramoth-Gilead. The stele is aimed at dissuading rebellion against him and association with Jehoash of Israel.
\nHowever, as Adad-Nirari III departs for Mesopotamia, Jehoash takes advantage of Bar Hadad\u2019s weakness and begins a campaign against him to recover lost territories. Jehoash defeats Bar Hadad at Apheq in the Jezreel Valley.
\n795\u2013792 BCE:
\nJehoash defeats Bar Hadad II twice more\u2014at Lo-Debar on the east bank of the Jordan, and at Qarn\u00e1yim on the King\u2019s Highway in Transjordan. As a result, Bar Hadad loses all control of Transjordan and Galilee, thus relinquishing the southern trade routes. A new era of ascendancy and prosperity begins for Israel.
\nc. 792\/91 BCE:
\nJehoash, now in possession of Dan, breaks down Bar Hadad\u2019s stele at the city gate. Some of the fragments are then recycled as building material in a new construction phase at the city gate.<\/p>\n
Bayt-Dawid and the Quest for King David<\/p>\n
Perhaps the most hotly contended of the many controversial issues about the Tel Dan Inscription has been the interpretation of the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Bayt-Dawid) in Line A9. It is certainly not surprising that the word became a battleground for ideologies on historical methodology, for the very use of the element \u05d3\u05d5\u05d3 raises issues about the historicity of David and the United Monarchy as well as the interaction between archaeology and biblical texts. The use of the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Tel Dan Inscription begs the question, \u2018does this refer to King David?\u2019 Indeed, scholarship must deal with such questions and others like it arising from the Tel Dan Inscription.
\nIn the textual analysis we saw how context and syntax demand that we consider Bayt-Dawid as a toponym. This is, however, at odds with the suggestion that the lexeme \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 is akin to such state-dynastic labels as B\u00eet \u1e2aumri. Rather than represent a political entity, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 signifies a geographical location\u2014namely, Jerusalem\u2014which was known as \u2018Bayt-Dawid\u2019. The present investigation now turns to the derivation of the individual element \u05d3\u05d5\u05d3 within the toponym \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3.
\nWe must realize, however, that we have a methodological dilemma as we approach this question. History is reconstructed through the use of numerous sources, each weighted according to their reliability and value. If we proceed to assess the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Tel Dan Inscription in accordance with all our previous data and conclusions, we risk making a premature judgment on the word rather than incorporating it into our data to give the evidence a fuller scope on which to base conclusions. That is, we run the risk of circular argumentation by using previously formed conclusions as evidence for conclusions themselves, thereby leaving the Tel Dan Inscription out of the loop of evidence. On the other hand, if we proceed to incorporate it as part of our data, we risk not having assessed its meaning and worth properly before using it for reconstruction, if indeed such assessment allows us to use the Tel Dan Inscription as part of the evidence.
\nNeedless to say, this dilemma exists with every new datum used for historical reconstruction. Yet there is a sharper edge to our dilemma with the word \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 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 \u05d3\u05d5\u05d3.
\nEtymologically and contextually, the Aramaic toponym \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Bayt-Dawid) is identical to the Hebrew toponym \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018Ir Dawid, or \u2018City of David\u2019). The biblical texts make the statement that Jerusalem was captured by David and renamed \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019) after himself (2 Sam. 5:7, 9; 1 Chron. 11:7). However, many, such as Davies, Lemche and Thompson, 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 \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019), and therefore its Aramaic equivalent \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (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.
\nThompson, however, does analyze the biblical texts. He does this in order to ascertain contextually how the name \u2018David\u2019 (\u05d3\u05d5\u05d3) was understood 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 earlier time. According to Thompson, the issue is one of priority: whether the name Bayt-Dawid (\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3) was derived from the name of a personage, \u2018David\u2019, or whether the name \u2018David\u2019 was derived from the word Bayt-Dawid (\u05d1\u05d9\u05ea\u05d3\u05d5\u05d3). If the latter were the case, then the name \u2018David\u2019 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.
\nAt this juncture it is of paramount importance to distinguish between the use of the term \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Aramaic Tel Dan Inscription and the term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 in the Hebrew narratives of biblical literature. Despite appearances, the Aramaic term is not the equivalent of the Hebrew term. In the Tel Dan Inscription, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 is a toponym whose Hebrew equivalent is \u05d3\u05d5\u05d3 \u05e2\u05d9\u05e8. The Hebrew term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 is not a toponym, but a socio-political label, \u2018House of David\u2019. Thus, in dealing with the usage of the Hebrew term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3, we are not explaining the usage of the Aramaic term \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Bayt-Dawid) within the Tel Dan Inscription. Rather, we are attempting to trace the etymology of the element \u05d3\u05d5\u05d3 to identify its referent. If we can trace the referent in the Hebrew label \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) we may be in a better position to determine the referent in the Aramaic toponym \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Bayt-Dawid) as it is seen in the Tel Dan Inscription.
\nIn his contextual analysis of the term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) in the Hebrew narratives, Thompson highlights the eponymic nature of the label. Correctly, he brings out the idea that the term \u05d1\u05d9\u05ea (\u2018house\u2019) is a socially loaded word denoting a patron relationship, be it of a social, political or theological nature. The nature of such patronage, however, is more complex than Thompson\u2019s assessment. Such patron labels utilize family terminology to describe the relationship, but different labels describe different types of patron relationships. Thus, there is a need to distinguish between patron relationships labelled as \u05d1\u05d9\u05ea (\u2018house\u2019) and other relationships labelled as \u05d1\u05df\/\u05d0\u05d1 (\u2018father\/son\u2019). Thompson sees in the Hebrew term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) a theological patronage\u2014namely, that of Yahweh for the kings of Jerusalem. That is, the kings of Jerusalem were in a bound relationship of subordination to their superior, the patron deity Yahweh.
\nHowever, an appreciation of the Semitic root \u221a\u05d3\u05d5\u05d3 as well as certain biblical texts highlight that this type of relationship between a divine patron and a subordinate human was not described by the label \u05d1\u05d9\u05ea (\u2018house\u2019). First, the root \u221a\u05d3\u05d5\u05d3 denotes a close, loyal relationship, with nuances including \u2018uncle\u2019, \u2018aunt\u2019, \u2018darling\u2019 and \u2018love\u2019. We cannot consider \u2018general\u2019 as a nuance since Tadmor put an end to that suggestion long ago by demonstrating that the word dawidum, known from Mari texts, was simply a dialectal form of the more normative dabdu, meaning \u2018defeat\u2019. To the other legitimate nuances mentioned above we should add \u2018patron\u2019. Thus, the use of the noun \u05d3\u05d5\u05d3\u05d4 in the Mesha Stele need not be as enigmatic as is often thought if we translate the noun as \u2018his patron\u2019. Not only is such a translation in keeping with the meaning of the Semitic root \u221a\u05d3\u05d5\u05d3, but it makes good contextual sense. A deity, therefore, could legitimately be given the title \u05d3\u05d5\u05d3. However, biblical texts demonstrate that it is also used as a human appellation, most notably in the personage of David (\u05d3\u05d5\u05d3 or \u05d3\u05d5\u05d9\u05d3). In this regard, the biblical name \u2018David\u2019 can be seen as a title that takes on the significance of a personal name, similar to an example cited by Thompson\u2014namely, \u2018Bar Hadad\u2019 derived from \u2018son of Hadad\u2019.
\nYet, Thompson argues that the biblical character David is the personification of Yahweh\u2019s patronage of Jerusalem\u2019s 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 personal name \u05d3\u05d5\u05d3 suggests that it arose not from the notion of Yahweh\u2019s patronage of Jerusalem\u2019s kings, but from these kings\u2019 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 \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) the element \u05d3\u05d5\u05d3 refers to the patriarch of an aristocratic family landed in Jerusalem (as well as Bethlehem and Hebron) who was acknowledged with the title \u05de\u05dc\u05da (\u2018king\u2019). In other words, the label \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) 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 \u2018subjects\u2019. The House of David is only seen to stand for Yahweh through the use of metonymy.
\nCertain key texts provide the evidence.<\/p>\n
2 Samuel 7:5\u201316. This text relates the divine oracle given to David about the establishment of David\u2019s house. The language is very clearly that of Yahweh\u2019s patronage of David and his descendants. Although the term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 in is never used, it is implied numerous times. Of particular importance is the consistent parallelism between the terms \u05d1\u05d9\u05ea (\u2018house\u2019) and \u05de\u05de\u05dc\u05db\u05d4 (\u2018kingdom\u2019). This indicates that the label \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 did not refer primarily to Yahweh\u2019s patronage of Jerusalem\u2019s kings, but to the kings\u2019 patronage of the town of Jerusalem. In distinction, Yahweh\u2019s divine patronage of Jerusalem\u2019s kings is described with different patronage terminology\u2014namely, \u05d0\u05d1 (\u2018father\u2019) and \u05d1\u05df (\u2018son\u2019). That is, the relationship between Yahweh and Jerusalem\u2019s kings was not one of extended family (\u05d1\u05d9\u05ea, \u2018house\u2019) which saw homage rendered to the \u2018first among equals\u2019 as it were, but of seniority in which an inferior (\u05d1\u05df, \u2018son\u2019) rendered homage to a superior (\u05d0\u05d1, \u2018father\u2019). This relationship of seniority is further demonstrated in David\u2019s response (2 Sam. 7:17\u201329) which describes the relationship as one between \u05d0\u05dc\u05d4\u05d9\u05dd (\u2018god\u2019) or \u05d0\u05d3\u05d5\u05df (\u2018master\u2019) and \u05e2\u05d1\u05d3 (\u2018servant\u2019). Thus, the term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 certainly describes patronage, but it is not that between Yahweh and the kings, nor even between Yahweh and his people.<\/p>\n
1 Kings 12:16\u201333. This famous text deals with the response of the Israelite tribes to Rehoboam\u2019s harsh demands, as well as Jeroboam\u2019s 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\u2019 lips, \u2018What share do we have in David?\u2019 (1 Kgs 12:16), makes little contextual sense. Indeed, Jeroboam\u2019s 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 Jerusalem, 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\u2019s kings, but as divine patron of all the Israelite tribes. However, the prime referent of the name \u05d3\u05d5\u05d3 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 \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) is still meaningful as a reference to the kings of Jerusalem, exemplified specifically here by Rehoboam.<\/p>\n
1 Kings 14:7\u201311. This text, representing a divine utterance of Yahweh to Jeroboam, is particularly important, for in it Yahweh portrays himself as Jeroboam\u2019s patron. This, however, does not make Jeroboam a member of the House of David. Rather, there is a deliberate contrast between the House of Jeroboam and the House of David (1 Kgs 14:8, 10). As such, the referent of \u05d3\u05d5\u05d3 (\u2018David\u2019) is not Yahweh as patron, but rather the kings of Jerusalem. More specifically, \u05d3\u05d5\u05d3 (\u2018David\u2019) is the king whom Yahweh 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 (Rehoboam) because he is earlier portrayed as a fool in order to accomplish Yahweh\u2019s turn of events (1 Kgs 12:15). Rather, \u05d3\u05d5\u05d3 must refer back to the eponymous ancestor of Jerusalem\u2019s king\u2014namely, David ben-Jesse.<\/p>\n
Jeremiah 21:11\u201312. This prophetic speech calls on the House of David, represented by the king of Judah, to fulfil the role of patron by administering justice to the oppressed. It is telling that this saying uses the patron label \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) since it is in the capacity of patron that the House of David is addressed. The inference is that the House of David\u2019s patronage reflects back on Yahweh, hence the threat of Yahweh\u2019s rage. However, there is a very clear difference between Yahweh and the House of David. In fact, the inference that Yahweh\u2019s reputation is being tarnished highlights that the connection between Yahweh and the House of David is one of contrast.<\/p>\n
2 Chronicles 21:7. The Chronicler\u2019s comment about Yahweh being unwilling to destroy the House of David sees a definite distinction between Yahweh and the House of David. There is no metonymic association between the two.<\/p>\n
These texts are merely a selection, demonstrating that any connection between Yahweh as patron deity and \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 (\u2018House of David\u2019) is purely metonymic. They are certainly not interchangeable labels and there is no reason to suggest the latter is derived from the former. On the contrary, the evidence demonstrates that the term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 specifies an aristocratic family as patrons of Jerusalem and the surrounding peoples. Thompson is close to the mark when he sees the term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3 representing a godfather, but the actual referent is the king of Jerusalem, not the deity Yahweh.
\nThough we can no longer see Yahweh as the referent in the biblical term \u05d1\u05d9\u05ea \u05d3\u05d5\u05d3, we must still ask whether the Aramaic term \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 originated as a reference to a temple. Numerous toponyms with the element \u05d1\u05d9\u05ea\u05be (\u2018House of \u2026\u2019) reflect the nuance of \u2018temple\u2019. This is seen in such toponyms as \u05d1\u05d9\u05ea\u05d0\u05dc (Bethel: \u2018House of El\u2019), \u05d1\u05d9\u05ea\u05be\u05e9\u05de\u05e9 (Beth-Shemesh: \u2018House of [the sun god] Shamash\u2019) and \u05d1\u05d9\u05ea\u05be\u05d7\u05d5\u05e8\u05df (Beth-\u1e24oron: \u2018House of \u1e24auron\u2019). However, \u05d1\u05d9\u05ea\u05be toponyms are not exclusively connected with deities. Some such toponyms are named after geographical features, such as \u05d1\u05d9\u05ea\u05be\u05d4\u05e2\u05e8\u05d1\u05d4 (Beth-Arabah: \u2018House of the Arabah\u2019) and \u05d1\u05d9\u05ea\u05be\u05d4\u05e2\u05de\u05e7 (Beth-Emeq: \u2018House of the Valley\u2019), while others are named after various attributes, such as \u05d1\u05d9\u05ea\u05be\u05e9\u05d0\u05df (Beth-Shean: \u2018House of Ease\u2019), \u05d1\u05d9\u05ea\u05be\u05e4\u05dc\u05d8 (Beth Pelet: \u2018House of Escape\u2019), \u05d1\u05d9\u05ea\u05be\u05d2\u05d3\u05e8 (Beth-Gader: \u2018House of a Wall\u2019) and \u05d1\u05d9\u05ea\u05be\u05d7\u05d2\u05dc\u05d4 (Beth-Hoglah: \u2018House of a Partridge\u2019). If \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 of the Tel Dan Inscription did originate with a temple of some kind, then we must interpret the toponym as meaning \u2018House of (the) Patron\u2019. This may or may not have referred to a deity. Thus, although the toponym \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in the Tel Dan Inscription can be connected indirectly with a deity, such a connection is not demanded by the toponym itself.
\nIt is not uncommon for cities in antiquity to be named after individuals. That this was a common phenomenon is evidenced by finds across the entire Near East. In the Inscription from Karatepe, Azitawadda names a city after himself, the city of Azitawaddiya. From Assyria, we have many cities named after kings, such as Kar-Tukulti-Ninurta, Kar-Shulmanasharidu and Dur-Sharrukin. That Jerusalem was named after David is a tradition preserved in biblical literature, and there is certainly nothing extraordinary about such a story. The only moot point is the existence of an historical David. This opens up two possibilities for us: Jerusalem may have been renamed after an historical personage named \u2018David\u2019, or it may have been renamed after a mythic eponymous ancestor given that name.
\nUndeniably, at a certain stage in history, Jerusalem was known as \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (Ir David: \u2018City of David\u2019). That the toponym \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 is the Aramaic equivalent for the same location has already been established. Thus, we need to understand how Jerusalem came to be given the name \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 or \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3. From the Amarna Letters, we know that the city was known as Urusalim (\u2018Jerusalem\u2019) during the Amarna Period (fourteenth century BCE). From Sennacherib\u2019s annals we know that by the late eighth century BCE, during the time of Hezekiah, the city was known by this name again. The Tel Dan Inscription, however, indicates that less than a century before Hezekiah, Jerusalem was known by another name, \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (Ir David: \u2018City of David\u2019).
\nThis interval between the Amarna Period and the time of Sennacherib (Late Bronze II-Iron II) was a veritable dark age as far as Jerusalem is concerned. The city seems to have declined during and shortly after the Amarna Period, presumably because of Egyptian hegemony and the social upheaval reflected in the Amarna correspondence. However, the town never appears to have been uninhabited during this time. Occupying the Ophel Ridge between the Qidron and Tyropoean Valleys, the remains of the site resemble a small fort rather than an actual town or village. Excavations 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 Jerusalem any sort of relevance on the political stage. Evidence suggests that Jerusalem, known at this time as \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (or \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 in Aramaic), was probably the home of an aristocratic family with a notorious reputation or past\u2014a kind of self-important \u2018mafia\u2019 clan.
\nIn comparison, the biblical record presents a very similar picture. Traditional notions down through the ages and into our time have built up fanciful pictures of Jerusalem which, on closer inspection, do not actually reflect the biblical picture itself. 1 Samuel 5:6\u20139 portrays David as living in his newly conquered compound and renaming the compound from \u05e6\u05d9\u05d5\u05df (\u2018Zion\u2019) to \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019). 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 population\u2014not 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, \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (or \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 in Hebrew).
\nFurthermore, it is noteworthy that in the book of Kings, the usage of the name \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019) 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\u2019s kings. Ahaz is the final king explicitly said to have been buried in \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3, the \u2018City of David\u2019 (2 Kgs 16:20), though the Chronicler may give slightly different information when he puts Ahaz\u2019s burial place \u2018in the city, in Jerusalem\u2019 (2 Chron. 28:27). Hezekiah\u2019s burial is glossed over without details in 2 Kgs 20:21, but the Chronicler tells us that he was buried \u2018in the upper of the tombs of David\u2019s sons\u2019 (2 Chron. 32:33). Presumably, this means Hezekiah was also buried in the royal compound known as \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019).
\nDuring the reigns of both Ahaz and Hezekiah (c. 732\u2013687 BCE), Jerusalem was expanding beyond the fortified compound and the entire Ophel Ridge. Jerusalem reaped the economic benefits of its expanding sovereignty 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\u2019s reign that the city was besieged by Sennacherib (701 BCE) who recalls the name of the city as Ursalimu (\u2018Jerusalem\u2019). 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 \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019) referred in the strictest sense to the fortified compound of Jerusalem\u2019s 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 \u2018Jerusalem\u2019 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 \u2018Jerusalem\u2019 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.
\nFurthermore, 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\u2013Kings, as well as the Chronicler, and the kings themselves. If many of Jerusalem\u2019s kings were entombed within the original royal compound\u2014the \u2018City of David\u2019\u2014then 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 numerous generations of Jerusalem\u2019s rulers. Although this probably does not completely bridge the gap between Jerusalem\u2019s kings and the biblical writers, it certainly closes it considerably.
\nThompson\u2019s assessment of the connection between referee and referent is, however, far more conservative:<\/p>\n
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\/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\u2019 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 \u2026<\/p>\n
Being reluctant to connect a source with a far distant referent seems, at first, to have a healthy dose of caution. However, the archaeological record, which becomes fuller with the Tel Dan Inscription, shows that such treatment of the biblical literature does not simply err on the side of caution, but denies the significance of very real connections between the kings of Israel and Judah in the biblical narrative and the appearance of these kings in the archaeological record. In terms of progression or chronology, there is little to poison our confidence in the catalogue of these kings. Some may question or doubt the historical events attributed to them, but the archaeological record affirms the existence of many of these kings. The evidence, then, suggests that the kings in the biblical catalogue did indeed exist, even if we do question what they did historically. In this vein, the mention of Jerusalem as \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (Bayt-Dawid) in the Aramaic Tel Dan Inscription overlaps considerably with the biblical Hebrew evidence, which sees Jerusalem referred to as \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 (\u2018City of David\u2019). And if, in this significant overlap of evidence, the fortress known as \u05e2\u05d9\u05e8 \u05d3\u05d5\u05d3 contained the remains of the majority of these kings, then our confidence can be boosted that the catalogue is accurate.
\nThompson may well be correct in his suggestion that the Hellenistic era is the milieu belying the biblical narratives of the kings. However, as the archaeological record becomes fuller, it becomes less likely that the referents in the narrative belong to the Hellenistic era. Circumstances in the Hellenistic era (or even the Persian era) may have been ripe enough to motivate the writing or compilation of the narrative, but this is altogether different from claiming that the referents in the narrative actually are from that era. Similarly, we must be careful not to confuse the referent of a narrative character with the function of that character within the narrative.
\nThe figure of Omri is a case in point. There is no concomitant reason to see the referent of Omri in the biblical narrative as being in the Hellenistic era, or whichever era the biblical narrative was written in. There are, however, very good reasons for seeing the referent as being the same as the entity to whom Mesha refers in his stele as Omri\u2014namely, a ninth century 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 referents\u2014one in ninth century BCE Moab and the other in second century BCE Judaea. Rather, the different functions and characteristics of each derive from the two different milieus that lay behind the writing of each text.
\nWhen it comes to the biblical figure of David, we can be confident that the referent is not an entity in the author\u2019s own era, nor a figure purely personifying Yahwistic patronage for the kings of Jerusalem. This is especially 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 biblical texts, rather than the imperial persona at the head of a great civilization 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 reasonable scenario.
\nHaving 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 \u05d1\u05d9\u05ea\u05d3\u05d5\u05d3 (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\u2019s 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 \u05d3\u05d5\u05d3 in its name, and Jerusalem is the logical identity of that location. To say that David is the original referent 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 Jerusalem. This, however, is because that naming event is outside the inscription\u2019s 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.<\/p>\n
Cultic Implications<\/p>\n
The Tel Dan Inscription alerts us to certain cultic developments that were occurring in Syria-Palestine at the time the inscription was produced\u2014that 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 of ma\u1e63\u1e63ebot, or sacred stones.
\nIn 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\u2019s treaty with Baal, king of Tyre (IV 6; c. 676 BCE), and in his Succession Treaty (line 467; c. 672 BCE). 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 \u05d0\u05dc (\u2018El\u2019) largely disappears as a theophoric element in Aramaic personal names after the seventh century BCE, replaced instead by \u05d1\u05d9\u05ea\u05d0\u05dc or \u05d1\u05ea\u05d0\u05dc (\u2018Baytel\u2019\/\u2018Bethel\u2019).
\nSince 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 \u2018El\u2019 to subsequent \u2018Bethel\u2019. The name \u05d0\u05dc\u05af\u05d1\u05d9[\u05ea\u05d0\u05dc] (\u2018El-Baytel\u2019) in Line A4 should be seen as a stage in the deity\u2019s 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 reliable, 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\u2019s existence in Palestine during and probably before the sixth century BCE.
\nThere is also good reason to suggest a connection between this deity and sacred Bethel-stones (or \u2018Baetyls\u2019) which were venerated at cultic shrines. The term \u2018Bethel\u2019 for such sacred stones is well known from both Aramaic and later Greek texts where it appears as an element in numerous compound names of deities. These include \u2018Anat-Bethel, Ishum-Bethel and \u1e24erem-Bethel in the Aramaic papyri of Elephantine, and Zeus Betylos, Zeus Baitylos and Symbetylos in later Greek inscriptions from Syria and Mesopotamia. Baitylos is also explicitly named as a \u2018stone-god\u2019. The fact that \u2018Anat 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 Palestine without compounds. Thus, the compound divine names may be understood as appositional constructions. So, \u2018Anat-Bethel means \u2018The Bethel-stone, \u2018Anat\u2019. In regard to the divine name \u2018El-Baytel\u2019 in the Tel Dan Inscription, we may understand this name in the same way, meaning \u2018The Bethel-stone, El\u2019.
\nSacred Bethel-stones are well attested in the Levant and in the Phoenician colonies. In all probability, the ma\u1e63\u1e63ebah uncovered at numerous Levantine sites, including Tel Dan, were venerated as sacred objects. The term \u05d1\u05d9\u05ea\u05d0\u05dc (\u2018house of god\u2019 or \u2018divine house\u2019) may have been a reference to the entire shrine in which a ma\u1e63\u1e63ebah stood. These ma\u1e63\u1e63ebah, 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 ma\u1e63\u1e63ebah 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.
\nThe Jacob narratives are particularly instructive here as to the nature and function of Bethel-stones or ma\u1e63\u1e63ebah. After his nocturnal vision of Yahweh at Luz, Jacob sets up the stone on which his head rested as a ma\u1e63\u1e63ebah. He then takes a vow concerning this ma\u1e63\u1e63ebah. The implication of this is that the ma\u1e63\u1e63ebah 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. Other texts, such as Gen. 31:13, Jer. 2:27, Dan. 2:34\u201335 and Zech. 3:9 also demonstrate how these sacred stones were considered animate objects.
\nIn this way, it is not surprising to find ma\u1e63\u1e63ebah in the ninth\u2013eighth century BCE gate complex of Tel Dan. Just inside the so-called \u2018Outer Gate\u2019, we find five ma\u1e63\u1e63ebah set up along a fortification wall to the immediate right of the gate\u2019s threshold. These ma\u1e63\u1e63ebah range in size from 30 cm to 50 cm tall. Next to them, in the destruction layer that covered them, the excavators uncovered a small assemblage of pottery, including incense bowls, oil lamps, bowls and plates. This confirms that the ma\u1e63\u1e63ebah were venerated as sacred stones or Bethel-stones. Their position in this gate chamber is interesting because of the discovery of a canopied platform in the same chamber. This platform would have borne the throne of either the king or a high official who acted as judge. In this way, the sacred stones were witnesses of the justice dispensed at the gate as well as of contracts and covenants concluded there. In all likelihood, the relevant parties of a dispute or contract made some kind of offering to the ma\u1e63\u1e63ebah deities. Since the deity El was widely viewed as both wise and just, able to judge disputes correctly, we may reasonably expect that he was represented by at least one sacred Bethel-stone or ma\u1e63\u1e63ebah at the city gate of Tel Dan and of other sites with similar constructions.
\nma\u1e63\u1e63ebah were also discovered at the base of Dan\u2019s city wall to the east of the gate chambers where justice was dispensed. One cluster of three such ma\u1e63\u1e63ebah was discovered at the rear of a rectangular stone construction with dimensions 4.5 \u00d7 2.5 m. This construction acted as a platform, raised slightly above the surrounding ground level. The largest ma\u1e63\u1e63ebah measured 1.17 m tall and a basalt bowl, with traces of fire in it, were found directly in front of it on a decorative capital. Much ash was also found around all the stones, confirming this as a cultic shrine at which offerings were made. I propose that the largest ma\u1e63\u1e63ebah was representative of El-Baytel and that the decorative capital may have been a sacred hearth (\u05d0\u05e8\u05d0\u05dc), or even that the entire platform was considered a sacred hearth. 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.
\nAlso interesting in explaining the function of sacred stones are the texts of Gen. 35:19\u201320 and 2 Sam. 18:18. In Gen. 35:19\u201320 we read that after Rachel\u2019s death, Jacob set up a ma\u1e63\u1e63ebah 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 ma\u1e63\u1e63ebah in order to preserve the memory of his name because he had no son. Interestingly, he is said to have named the ma\u1e63\u1e63ebah after himself. This is particularly noteworthy for the Tel Dan Inscription. In wishing that his father be remembered \u2018at every ancient hearth on[sacred] ground of El-Baytel\u2019, 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\u2019s 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 \u2018joined\u2019 El in the afterlife. It seems plausible that this belief was carried on considering the connection between El and El-Baytel.
\nAlternatively, we may understand Bar Hadad\u2019s 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.
\nThe 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\u2013eighth centuries BCE). This cultic platform measures 2.31 \u00d7 1.53 m at the base. Two steps, however, lead up to a raised square platform that is 1.53 \u00d7 1.53 m in area. The platform was 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 (\u05d0\u05e8\u05d0\u05dc). It is also evident from a depression at the far end of the platform that the basin stood before a ma\u1e63\u1e63ebah. A decorated iconic ma\u1e63\u1e63ebah found nearby had been deliberately knocked down and broken into five pieces. The relief on this iconic ma\u1e63\u1e63ebah has been interpreted as an Aramaean manifestation of the moon god. However, it is debatable whether this ma\u1e63\u1e63ebah was the one originally located on the platform. Whatever the case, it is evident that these ma\u1e63\u1e63ebah were associated with legal proceedings at the city gate, as well as the memory of prominent deceased persons.
\nThe fact that the name of \u2018El\u2019 is largely replaced by \u2018Baytel\u2019 or \u2018Bethel\u2019 from the seventh century BCE onwards suggests that El came to be known as \u2018El-Baytel\u2019, meaning \u2018The Bethel-stone, El\u2019, prior to the seventh century. 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 complexes, 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.
\nThis personification of El as a sacred Bethel-stone, presumably the chief Bethel-stone at the city gate, probably led to the subsequent apocopation of \u2018El-Baytel\u2019 to simply \u2018Baytel\u2019. Thus, from the seventh century BCE onwards, reference to the deity \u2018Baytel\u2019 or \u2018Bethel\u2019 without any prefixed compound element should be understood as a reference to the deity El in his evolved state. Other deities whose names are compounded with \u05d1\u05d9\u05ea\u05d0\u05dc (\u2018Bethel\u2019) may, however, be understood as distinct from the deity Bethel, their names being appositional constructions as mentioned previously.
\nThus, Esarhaddon\u2019s call upon Bethel and \u2018Anat-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 \u2018Anat in their representations as Bethel-stones and witnesses to the treaty, to deliver the defaulting parties \u2018to the paws of a man-eating lion\u2019. From Elephantine, the letter of Yedonyah and his colleagues to Bigwai, governor of Judaea, informs us that the temple at Elephantine had stone pillars. 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 \u2018the pillars of stone\u2019 (\u05e2\u05de\u05d5\u05d3\u05d9\u05d0 \u05d6\u05d9 \u05d0\u05d1\u05e0\u05d0) in the temple were sacred Bethel-stones representing the five deities mentioned specifically as revered by the Judaean community in Elephantine.
\nIn 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 of ma\u1e63\u1e63ebah, especially those found at various Iron Age gate complexes.<\/p>\n
Chapter 8<\/p>\n
CONCLUDING REMARKS<\/p>\n
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 demise of 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 successful father, Hazael. Bar Hadad\u2019s 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\u2019 diminishing power.
\nThe inscription also further defines our picture of ancient Jerusalem. In c. 800 BCE Jerusalem is seen to have been little more than a fortified 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 cannot 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.
\nEpigraphically, 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 B1 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.
\nWith regards to the biblical texts, the Tel Dan Inscription demonstrates that there are definite historical kernels in the Bible that cannot readily be dismissed. Indeed, the biblical records are an integral part in reconstructing the history of Syria-Palestine. However, these need to be first evaluated in the light of archaeological discoveries, such as the Tel Dan Inscription. The method that should be followed is to examine inscriptions independently of biblical texts in the first stage. In this way, the epigraphical, palaeographical and philological considerations may proceed without external influence. Once the inscription has been appraised, we may corroborate it with the biblical texts.
\nWhat this process shows is that there are definite historical memories in the biblical literature. However, these memories have developed. With the light that archaeology sheds on the biblical literature we can see something of how these memories have developed and been embedded in the biblical text. Since such archaeological finds as the Tel Dan Inscription are rare, we do not have clarity for all the history of Syria-Palestine in the Iron Age. However, with more similar discoveries, our image of the history will hopefully sharpen.
\nThe importance of archaeology is also underlined when we consider the issue of dating the Tel Dan Inscription. The archaeological context is perhaps the most important factor here. Dates determined on purely epigraphical and palaeographical avenues are far from being reliable and need to be corroborated with the archaeological context of the inscription before they can be useful. In the case of the Tel Dan Inscription, all these factors worked together and confirmed a date for the inscription at c. 800 BCE, or more specifically, in the early eighth century BCE. It was then with other evidence, such as the Assyrian and biblical texts, that we could refine this date even further. These methods allowed us to determine that the Tel Dan Inscription had a display life between c. 796\u2013791 BCE before it was broken down and recycled as building material during the Israelite renaissance at Dan. The value of archaeology cannot be denied.
\nThe Tel Dan Inscription also brings us a definite step closer to finding a historical David. This was perhaps the most hotly contested issue among scholars when the fragments were first discovered. The reappraisal and subsequent new reading and interpretation offered in this study have not settled the issue beyond all doubt. The evidence proceeding from the inscription cannot be construed as proof. Rather, the inscription has increased the likelihood of a historical David.
\nThe Tel Dan Inscription and the questions it raises must warn scholars of a future generation about the process of drawing historical conclusions from archaeological artefacts. It is not surprising that an epigraphical find 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.
\nThe primary lessons that we must learn are with regard to the investigative process itself. First, although resources are often frustratingly limited, scholars must not be tempted to make conclusions over inscriptional artefacts based on photographic or hand-drawn images while the actual inscriptional remains are still extant. As was the case with the Mesha Inscription, there are occasions when scholars cannot rely on anything but manufactured 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 thorough (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.
\nSecond, 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 Jerusalem, if indeed it is even a question that should be asked. The extensive conclusions of Jamieson-Drake 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 \u2018United Monarchy\u2019 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 themselves 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 \u2018United Monarchy\u2019 terms such as \u2018monarchy\u2019, \u2018state\u2019 or \u2018empire\u2019 are false or anachronistic at best. It may be the case that a period of United Monarchy\u2019 should be rebadged as the \u2018United Quasi-State\u2019 or \u2018United Para-State\u2019 or even \u2018United Racket\u2019. The answers to such questions must await further research. However, when interpreting archaeological remains and then comparing them with biblical texts, archaeologists and historians alike must be careful not to be comparing the archaeological record with false expectations or misunderstandings. In other words, it is a close and accurate reading of the biblical texts that must work alongside (not in replacement of) the interpretation of artefactual evidence. When it comes to the books of Samuel-Kings, we must look again at what the literature itself is saying.
\nIt is doubtless that discussion over the Tel Dan fragments will continue. It is envisaged, however, that this study will do much to quell the unhelpful passion and euphoria that the Tel Dan Inscription has evoked among scholars and interested persons alike. Although the Tel Dan fragments bring great excitement to the field of Biblical and Ancient Near Eastern Studies, we must not be lured into pursuing emotional scholarship. Rationality must prevail. It is also hoped that the conclusions from this study will generate a rational interest in the inscription, the history of Syria-Palestine in the Iron Age, as well as the study of biblical texts.<\/p>\n
series Journal for the Study of the Old Testament Supplement Series,
\ntitle The Tel Dan inscription: a reappraisal and a new interpretation,<\/p>\n","protected":false},"excerpt":{"rendered":"
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 … <\/p>\n