It is often assumed that early alphabetic writing was developed by members of a Semitic-speaking, Western Asiatic population (‘Canaanites’) who were involved in Egyptian mining operations around Serabit el-Khadim in the Sinai Peninsula (Sass Reference Sass1988; Goldwasser Reference Goldwasser2006; Naʾaman Reference Naʾaman2020). Later, this early alphabet would spread to the Southern Levant, where it was transformed into the Phoenician alphabet, from which the Greek alphabet subsequently derived (Albright Reference Albright1969; Naveh Reference Naveh1987; Sass Reference Sass1988; Goldwasser Reference Goldwasser2006; Hamilton Reference Hamilton2006, Reference Hamilton, Hackett and Aufrecht2014; Morenz Reference Morenz2011; Daniels Reference Daniels2017; Burlingame Reference Burlingame2019; Naʾaman Reference Naʾaman2020). This interpretation builds upon the discovery of a number of early alphabetic inscriptions that were discovered in Sinai from the early twentieth century AD onwards (Petrie Reference Petrie1906; Leibovitch Reference Leibovitch1934), at and around the temple of Hathor at Serabit el-Khadim (Figure 1). In an influential article, Alan Gardiner demonstrated that these inscriptions were examples of early alphabetic writing derived from Egyptian hieroglyphs (Gardiner Reference Gardiner1916; see also Gardiner Reference Gardiner1962).
In 1998, two further early alphabetic inscriptions were discovered in the Wadi el-Hol, in the western desert of Egypt, and tentatively dated to the late Middle Kingdom (late Twelfth to early Thirteenth Dynasties; late nineteenth/early eighteenth century BC) (Darnell et al. Reference Darnell, Dobbs-Allsopp, Lundberg, McCarter, Zuckerman and Manassa2005). These inscriptions demonstrate that early alphabetic writing was not confined to Sinai, but was also used in the Nile Valley. While most scholars agree that these inscriptions are indeed examples of early alphabetic writing (Hamilton Reference Hamilton, Hackett and Aufrecht2014: 31–34)—perhaps using Egyptian vocabulary (Wimmer & Wimmer-Dweikat Reference Wimmer and Wimmer-Dweikat2001)—others have suggested that they can also be understood as ‘proper’ Egyptian hieroglyphic inscriptions (Morenz Reference Morenz2011: 171–75). Another example of probable early alphabetic writing has been recently discovered on an ostracon (a potsherd used as a writing surface) from a tomb in western Thebes (TT 99) and dated palaeographically to the fifteenth century BC (Haring Reference Haring2015; Fischer-Elfert & Krebernik Reference Fischer-Elfert and Krebernik2016; Schneider Reference Schneider2018).
The date for the development of early alphabetic script is still disputed (Lemaire Reference Lemaire, Halpern and Sacks2017; Haring Reference Haring, Boyes and Steele2020). It was originally dated by Gardiner (Reference Gardiner1916: 13–14) to the Twelfth Dynasty (early second millennium BC)—a period of intensive Egyptian mining activity around Serabit el-Khadim. A date after 1500 BC (in the Eighteenth Dynasty), or even later, was soon suggested (Leibovitch Reference Leibovitch1934, Reference Leibovitch1963) and accepted by several other scholars (e.g. Albright Reference Albright1948, Reference Albright1969; Naveh Reference Naveh1987). Lemaire (Reference Lemaire, Halpern and Sacks2017: 106) suggests that early alphabetic writing was developed during the “period of Hyksos domination in the south of Palestine or in the Egyptian Delta around the 18th–17th century BCE”. Benjamin Sass (Reference Sass1988), who in the late 1980s advocated an early date, withdrew his original position in the early 2000s, arguing that “the alphabet was born in the 14th or early 13th century [. . .] surfacing in the Levant shortly thereafter” (Sass Reference Sass2004–2005: 157).
Farther north, the earliest occurrence of early alphabetic writing in the Southern Levant is also disputed. Scholars long assumed that the script was introduced to the region as early as the transition from the Middle to the Late Bronze Age (c. 1600/1550 BC), on the basis of several inscribed objects (see below for details of the controversy) (e.g. Naveh Reference Naveh1987; Sass Reference Sass1988). Sass later (Reference Sass2004–2005: 157), however, argued that “no pre-14th-century, perhaps no pre-1300 BC alphabetic inscriptions from Palestine can be pointed out with any confidence”. Naʾaman (Reference Naʾaman2020) also recently dated the spread of alphabetic writing to the Southern Levant to not before Late Bronze II (fourteenth century BC), and argued that its development in the Southern Levant was linked to scribal activities in the Egyptian centres of the Late Bronze Age.
If we accept an early Twelfth Dynasty date for the inscriptions in Sinai, and Twelfth to early Thirteenth Dynasty dates for those from Wadi el-Hol, then this leaves a significant temporal gap between these inscriptions and the first securely dated early alphabetic inscriptions in the Southern Levant in the fourteenth/thirteenth centuries BC. A recently discovered early alphabetic inscription from Tel Lachish now fills this gap and sheds new light on disputed examples of early alphabetic writing that have previously been dated to the late Middle Bronze Age.
Archaeological context and dating
Tel Lachish, located in the Shephelah region in modern-day Israel, is one of the most prominent Bronze and Iron Age sites of the Southern Levant. Excavations at the site have unearthed substantial Late Bronze Age (c. 1600/1550–1200 BC) and Iron Age (c. 1200–586 BC) remains, yielding imports from Egypt, Cyprus and the Aegean that attest to the regional importance of the site. Late Bronze Age Lachish is also mentioned in Egyptian sources, such as Papyrus Hermitage 1116A from the time of Amenhotep II (c. 1427–1401 BC), which mentions an envoy from Lachish coming to the Egyptian court (Epstein Reference Epstein1963), and figures prominently in the Amarna letters, the cuneiform correspondence between the major powers of the time found at Tell el-Amarna in Egypt (Rainey Reference Rainey2015). An Austrian team renewed excavations at Tel Lachish in 2017 (Streit et al. Reference Streit, Webster, Becker, Jeske, Misgav and Höflmayer2018). Two excavation areas are currently under investigation (Figure 2): area P, located to the north of the Iron Age Judean Palace-Fort, and area S at the western slope of the mound. Both of these areas were originally excavated in the 1970s and 1980s by the Tel Aviv University expedition (Ussishkin Reference Ussishkin2004a), and have now been re-opened in order to deepen our understanding of the Middle and Late Bronze Age strata, with the aid of high-resolution AMS radiocarbon-dating.
The new early alphabetic inscription was found during the 2018 excavation season in area S, in locus L1114 of stratum S-3b. This area was originally excavated between 1973 and 1987 by the Tel Aviv team, who uncovered a sequence of Iron Age and Late Bronze Age occupation levels (Table 1) (Barkay & Ussishkin Reference Barkay, Ussishkin and Ussishkin2004a & Reference Barkay, Ussishkin and Ussishkinb).
Stratum S-3 is dominated by a monumental structure (building 100), with 1m-wide mud-brick walls set on unusually deep stone foundations. The southern wall (L1027) runs for 16m east–west, although most of the building lies beyond the northern section of the trench. Building 100 attests several phases of re-use, the latest of which was excavated by the Tel Aviv team (Barkay & Ussishkin Reference Barkay, Ussishkin and Ussishkin2004b: fig. 8.13). The Austrian expedition labels this last use phase as stratum S-3a, as two earlier sub-phases have now been identified: strata S-3b and S-3c. During each sub-phase, small domestic units were built immediately to the south of building 100, abutting wall L1027 (Figure 3). Excavations in 2019 revealed that building 100 was not an isolated monumental structure, but rather formed an integral part of what can be understood as a fortification system, including a city wall (L1220) and tower (L1127/L1163/L1227) (Figure 3).
The early alphabetic inscription (‘B10969’ on Figure 4) was found during excavation of deposits immediately to the south of wall L1027, close to the corner with city wall L1220, at a level (255m asl) near to its (wall L1220) uppermost preserved stones (this was prior to recognition and exposure of the stratum S-3c ‘fortification’). The inscription was undoubtedly deposited within the stratum S-3 phase, as it was sealed 0.5m below two stratum S-3a walls (L1162 and L1043) and the fragmentary stratum S-3a surface that was removed by the earlier Tel Aviv excavations (Barkay & Ussishkin Reference Barkay, Ussishkin and Ussishkin2004b: fig. 8.13).
The inscribed sherd was found at the boundary of three distinct loci (L1150, L1117 and L1114—a fill that overlies the other two), while exposing and sampling burnt lens L1150 for radiocarbon-dating. Although it is uncertain whether the sherd was deposited together with the organic material in lens L1150, it was certainly in direct contact with the burnt layer, and was therefore deposited at essentially the same time. Thus, the date of deposition of the sherd can be determined using radiocarbon-dating. Two measurements on separate barley grains from L1150 were analysed at the Groningen and ETH Zürich AMS laboratories (GrM-17566 and ETH-96577). The results are plotted in Figure 5, with both falling definitively in the fifteenth century BC, and most likely before its final quarter.
Over the past three years of excavation in area S, a detailed and robust radiocarbon sequence has been developed for strata S-3 and S-2, using short-lived material from a long series of in situ burnt layers (Webster et al. Reference Webster, Streit, Dee, Hajdas and Höflmayer2019). Bayesian modelling shows that the earliest stratum (S-3a) exposed by the Tel Aviv excavations dates to the fifteenth century BC, approximately 100 years earlier than previously thought (late fourteenth century BC; Ussishkin Reference Barkay, Ussishkin and Ussishkin2004b; Yannai Reference Yannai and Ussishkin2004). Figure 6 summarises the relevant outcomes of the most current Bayesian model, which includes 27 measurements and more than 15 sequential layers within the fine stratigraphy of strata S-2–3. Calculated transition boundaries are plotted, rather than individual dates, with the exception of GrM-17566 and ETH-96577, which have been included to highlight the constraining effect on L1150 (see Figure 6). Stratum S-3b and the find context of the inscribed sherd are constrained close to the mid fifteenth century BC (1460–1430 BC, 1σ; 1485–1425 BC, 2σ), providing a remarkably precise terminus ante quem for the inscription. The newly identified stratum S-3c—perhaps the phase from which the inscription itself originates—dates to the first half of the fifteenth century BC, which is contemporaneous with the city wall and early life of building 100.
Inscription and reading
The inscribed sherd is an approximately 40 × 35mm rim fragment from a Cypriote White Slip II milk bowl (Figure 7; Åström Reference Åström1972: 447–57). This ware first appeared during Late Bronze IB (e.g. Gittlen Reference Gittlen1981: 50–51), although it only became popular during Late Bronze IIA (Artzy Reference Artzy and Gitin2019: 342). At Tel Lachish, White Slip II ware was encountered in Fosse temple I (Tufnell et al. Reference Tufnell, Inge and Harding1940: 83, pl. XLIII B; Singer-Avitz Reference Singer-Avitz and Ussishkin2004), and appears in substantial quantities from Strata S-3 onwards (Bunimovitz Reference Bunimovitz and Ussishkin2004).
The inner surface of the sherd’s rim is inscribed in dark ink, with letters written diagonally. Two lines each containing three letters can be discerned. Two additional characters are visible on the right side of the upper line, and another is visible between the two lines. Our suggested reading for line one (the top line) is from right to left. The first letter can be identified as ʿayin (ע), which is based on the Egyptian hieroglyph ‘eye’ (Gardiner Sign List D4; Sass Reference Sass1988: 126–27; Hamilton Reference Hamilton2006: 180–88; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 131–32). As in most early alphabetic inscriptions from the Southern Levant, the letter is shaped like a circle, resembling an iris with the pupil missing. The second letter can be identified as bet (ב), which is based on the Egyptian hieroglyph ‘house’ (Gardiner Sign List O1; Sass Reference Sass1988: 111–12; Hamilton Reference Hamilton2006: 38–52; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 129). The letter has a rectangular shape with one corner open. The third letter can be read as dalet (ד), based on the Egyptian hieroglyph ‘door’ (Gardiner Sign List O31; Hamilton Reference Hamilton2006: 61–75; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 129). The suggested reading for this line may therefore be עבד, meaning ‘slave’, and could be part of a personal name. Names with the component ʿbd (slave) are very common in all Semitic languages, generally with a theophoric (bearing the name of a god) element of a local divinity (e.g. Tigay Reference Tigay1986; Golub Reference Golub2020).
The suggested reading for line two is also from right to left. The first letter can be identified as nun (נ), which derives from the Egyptian hieroglyphs ‘horned viper’ and/or ‘cobra’ (Gardiner Sign List I9 and I10; Sass Reference Sass1988: 125–26; Hamilton Reference Hamilton2006: 154–71; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 131). This letter can also be identified between lines one and two, and on the right side of line one. The second letter can be identified as pe (פ). While it is not entirely certain from which sign this character is derived, the hieroglyph for ‘corner’ (Gardiner Sign List O38) has been suggested (Sass Reference Sass1988: 128; Hamilton Reference Hamilton2006: 188–96). Goldwasser (Reference Goldwasser, Finkelstein, Robin and Römer2016: 132) argues that this sign is uncommon in Middle Kingdom inscriptions from Sinai, instead suggesting that the sign represents a building tool (Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 132). The third letter can be identified as tav (ת), and it is again unclear on which sign the character is based. It could be the hieroglyph for ‘crossed planks’ (Gardiner Sign List Z11; Hamilton Reference Hamilton2006: 246–53), but some have also suggested an independent origin (Sass Reference Sass1988: 133; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 134). The suggested reading for line two is therefore נפת, which in Hebrew means ‘honey’ or ‘nectar’. If read from left to right—תפנ—this term could be a verb from the root פני (‘to turn’), or part of an unknown name.
From a typological perspective, the letters seem to be of a somewhat later date to those from the inscriptions from Serabit el-Khadim, as suggested by the circle-shaped eye without the pupil and the relatively developed letter pe. Yet the letters of this inscription seem to be earlier than those from the later Late Bronze Age, like the Lachish Ewer (see below). A date at the beginning of the Late Bronze Age therefore seems reasonable, and is supported by the radiocarbon data mentioned above.
The newly discovered inscription from Tel Lachish is currently the earliest securely dated example of early alphabetic writing in the Southern Levant. In order to assess the importance of this find, we briefly review the other potential early alphabetic examples from the area.
A disputed contender for the earliest example is a scarab from Tell Abu Zureiq, in the Jezreel Valley. Found in a Middle Bronze Age tomb excavated by Meyerhof (Reference Meyerhof1989), the scarab was dated to the Thirteenth to Fifteenth Dynasties (Giveon Reference Giveon1988: 22; Keel Reference Keel1997: 17). Its base depicts a man and four signs, which Giveon (Reference Giveon1988: 22) originally interpreted as Egyptian hieroglyphs. Kitchen (Reference Kitchen1989) suggested that these signs could be read as early alphabetic characters, an interpretation rejected by Keel (Reference Keel1997: 16–17), but recently endorsed by Morenz (Reference Morenz2011: 164–65).
Another potential early alphabetic inscription is the much-discussed Lachish Dagger, which was discovered in 1934 by the British Expedition in tomb 1502, and dated to the late Middle Bronze Age (Tufnell Reference Tufnell1958: 254). The bronze dagger exhibits four potential early alphabetic signs (Tufnell Reference Tufnell1958: 128; Sass Reference Sass1988: 53–54; Hamilton Reference Hamilton2006: 390–91), and most scholars accept this interpretation (e.g. Albright Reference Albright1948, Reference Albright1969: 10; Naveh Reference Naveh1987: 26; Hamilton Reference Hamilton2006: 303–4; Goldwasser Reference Goldwasser2006: 132, Reference Goldwasser, Finkelstein, Robin and Römer2016: 140–42; Morenz Reference Morenz2011: 170–71; Lemaire Reference Lemaire, Halpern and Sacks2017: 106; Haring Reference Haring, Boyes and Steele2020: 59). In 1988, Sass agreed that the inscription was probably early alphabetic, pointing out that it would be the only one that could be securely dated to the Middle Bronze Age (Sass Reference Sass1988: 54). He later grew more cautious, however, and suggested that the signs might not be early alphabetic after all (Sass Reference Sass2004–2005: 150).
A third example that has been dated to the Middle Bronze Age is the so-called ‘Gezer Sherd’. Exhibiting three early alphabetic characters, this sherd was found in 1929 on the surface of Tel Gezer (Albright Reference Albright1935). It was soon dated to the Middle Bronze Age (Albright Reference Albright1935)—an attribution accepted by many scholars (e.g. Albright Reference Albright1969: 10; Naveh Reference Naveh1987: 26; Hamilton Reference Hamilton2006: 308–309; Morenz Reference Morenz2011: 166; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 143). Sass was more cautious, however, arguing that the sherd could not be classified typologically, and that its date could range from Middle Bronze Age to the Early Iron Age (Sass Reference Sass1988: 55). He later concluded that the Gezer Sherd is essentially undatable (Sass Reference Sass2004–2005: 149).
Several inscriptions on an assemblage of storage jars from Tel Gezer have also been interpreted as early alphabetic writing (Seger Reference Seger, Meyers and O’Conner1983, Reference Seger2013: 186–96; Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 142–43). These jars were found in storerooms next to the southern gate area (field IV) and were associated with stratum XVIII (early Late Bronze Age) and stratum XIX (late Middle Bronze Age) (Seger Reference Seger2013). Most of the jars were inscribed with a single sign, with only two jars bearing two signs each. Sass (Reference Sass1988: 98) mentioned these Gezer jars briefly as examples of early alphabetic writing, but later re-interpreted them as bearing only potters’ marks (Sass Reference Sass2004–2005: 166, footnote 97).
A fragmentary plaque from Shechem is frequently mentioned in the corpus of potential Middle Bronze Age early alphabetic inscriptions from the Southern Levant (Böhl Reference Böhl1938). According to the earliest publications, this object was found in a Middle Bronze Age building, just above the floor, together with typical, contemporaneous Tell el-Yahudiyah pottery (Böhl Reference Böhl1938: 2). Scholars have long accepted a Middle Bronze or early Late Bronze Age date for the plaque (Albright Reference Albright1948, Reference Albright1969: 10–11; Leibovitch Reference Leibovitch1963; Wimmer Reference Wimmer2001; Hamilton Reference Hamilton2006: 308), which represents the lower right portion of a stela depicting a person facing to the left and clad in a heavy garment (‘Wulstsaummantel’)—a common Middle Bronze Age garment type (Wimmer Reference Wimmer2001). The plaque’s archaeological context, however, has been questioned due to the early excavation techniques with limited stratigraphic control, and the lack of a final excavation report (Sass Reference Sass1988: 57). The early alphabetic nature of the characters has also been called into question (Sass Reference Sass2004–2005: 149–50).
Yet another disputed early alphabetic inscription was found at Tel Nagila in the 1960s. Here, a body sherd of a jug, with an inscription incised before firing, was discovered in area A, a residential area provisionally dated to the end of the Middle or the early Late Bronze Age (Amiran & Eitan Reference Amiran and Eitan1965: 121). Sass (Reference Sass1988: 54), however, rightly emphasised the lack of a clear stratigraphic context for that sherd. Later, quoting David Ilan, who observed that a large Late Bronze Age building disturbed the Middle Bronze Age strata in the area where the inscription was found, Sass concluded that the Tel Nagila sherd “is to be regarded as unstratified, and a LBII origin [is] not implausible” (Sass Reference Sass2004–2005: 159).
The dates and interpretations of the evidence for the earliest occurrences of early alphabetic writing in the Southern Levant are therefore ambiguous, as only the Lachish Dagger (if accepted as early alphabetic) was found in a clear archaeological context datable to the Middle Bronze Age (as rightly pointed out by Sass (Reference Sass1988: 54)). The discovery of the new early alphabetic inscription at Tel Lachish pushes back the earliest securely datable occurrence considerably, and we can now show that early alphabetic writing was employed in the Southern Levant by the mid fifteenth century BC (early Late Bronze Age). This evidence not only closes the gap between the development of early alphabetic inscriptions around Serabit el-Khadim and Wadi el-Hol in Upper Egypt, and its more widespread Southern Levantine use in the later Late Bronze Age, but also suggests that early alphabetic writing was already present in the Southern Levant by the (late) Middle Bronze Age.
The new early alphabetic inscription also underscores the importance of Tel Lachish as an early centre of writing (Goldwasser Reference Goldwasser, Finkelstein, Robin and Römer2016: 151; Naʾaman Reference Naʾaman2020). Indeed, Lachish has yielded more examples of Late Bronze Age early alphabetic inscriptions than any other site. In addition to the Lachish Dagger and the new inscription discussed here, the site has yielded four other examples of alphabetic writing. In tomb 527, the British Expedition of 1935 found a bowl (Lachish bowl one) bearing a painted inscription (Tufnell Reference Tufnell1958: 129). This tomb also contained a Cypriot Base Ring II juglet and a local imitation of a Mycenaean straight-sided alabastron (Tufnell Reference Tufnell1958: 239). Tufnell (Reference Tufnell1958: 129) considered this tomb to be contemporaneous with the late Fosse temple II or early Fosse temple III, and thus coeval (or slightly earlier) with stratum VII on the mound. In absolute terms, this dates to the fourteenth or thirteenth century BC (Ussishkin Reference Artzy and Gitin2004b: 57). In Fosse temple III, the British Expedition found the well-known Lachish Ewer, which bears a painted early alphabetic inscription (Tufnell et al. Reference Tufnell, Inge and Harding1940: 47–54; Tuffnell Reference Tufnell1958: 130). As Fosse temple III corresponds to stratum VII on the mound, the Ewer roughly dates to the thirteenth century BC (Ussishkin Reference Ussishkin and Ussishkin2004b: 57).
A fragment of a bowl bearing a black-ink inscription comprising two straight lines of characters was found by the Tel Aviv Expedition in pit 3867, in area S (Lemaire Reference Lemaire and Ussishkin2004). This pit belongs to stratum VI and dates to the twelfth century BC (Ussishkin Reference Ussishkin and Ussishkin2004b: 57). Finally, another inscription from stratum VI—a pottery sherd with several characters incised before firing—was found in the inner part of a Late Bronze Age temple in area BB during recent excavations by the Hebrew University of Jerusalem and Southern Adventist University (Sass et al. Reference Sass, Garfinkel, Hasel and Klingbeil2015).
Ben Haring (Reference Haring, Boyes and Steele2020: 62) has recently pointed out that many of the examples of early alphabetic writing “lack clear datings”, and thus our understanding of such writing before the well-known thirteenth century BC and later is limited. The new ostracon from Tel Lachish fills the gap between the potential early alphabetic writing on the late Middle Bronze Age Lachish Dagger and the corpus from the later Late Bronze Age phases.
The early alphabet developed in association with Western Asiatic (Canaanite) miners in Sinai (or, at least, was taken up by them) during the Middle Kingdom in the eighteenth century BC. We suggest that early alphabetic writing spread to the Southern Levant during the late Middle Bronze Age (with the Lachish Dagger probably being the earliest attested example), and was in use by at least the mid fifteenth century BC at Tel Lachish. Thus, the proliferation into the Southern Levant probably happened during the (late) Middle Bronze Age and the Egyptian Second Intermediate Period, when a Dynasty of Western Asiatic origin (the Hyksos) ruled the northern parts of Egypt. The new early alphabetic inscription from Tel Lachish provides fresh evidence to contextualise the spread of the early alphabet within the period of Hyksos domination over the Nile Delta and its still enigmatic connections with Middle Bronze Age city-states in the Southern Levant (cf. Lemaire Reference Lemaire, Halpern and Sacks2017). Furthermore, the new early alphabetic inscription dates to a period that also saw the earliest attested hieratic writing at Tel Lachish (Sweeney Reference Sweeney and Ussishkin2004: 1610–11), and when Lachish is mentioned for the first time in Egyptian sources during the reign of Amenhotep II (c. 1427–1401 BC) (Papyrus Hermitage 1116A; Epstein Reference Epstein1963; Webster et al. Reference Webster, Streit, Dee, Hajdas and Höflmayer2019). We now can show that early alphabetic writing in the Southern Levant developed independently of, and well before, the Egyptian domination and floruit of hieratic writing during the Nineteenth and Twentieth Dynasties (the thirteenth and twelfth centuries BC) (contra Naʾaman Reference Naʾaman2020).