The Tel Lachish script

The commonly accepted theory is that the Phoenician (Canaanite) alphabet was derived by reshaping Egyptian hieroglyphic signs over time through intermediate Proto-Sinaitic and Proto-Canaanite forms (Gardiner, 1916). Strictly, Proto-Sinaitic refers to the few inscriptions found in the Sinai Peninsula, hence the name. According to various authors, these are dated between the 18th and 16th centuries BC. Two lines of curious glyphs from the Wadi el-Hol valley, west bank of Nile Northwest of Luxor (Egypt), have been abusively added to the Proto-Sinaitic corpus. These are dated to the 19th - 18th century BC. Except for crosses and zigzag lines, the Wadi el-Hol and Sinai inscriptions have no signs in common. The Proto-Sinaitic signs resemble more Egyptian and Cretan hieroglyphics than Phoenician letters (see section The Proto-Sinaitic script). However, some authors believe them to represent alphabetic, phonetic writing attempts. Proto-Canaanite refers to later inscriptions found in Canaan and dated between 1450 and 1050 BC when the establishment of the classical Phoenician alphabet was conventionally fixed. The earliest Proto-Canaanite inscriptions are dated to Israeli and Levantine archeological age LB-II (late Bronze Age period II), i.e., 1400 – 1200 BC (Finkelstein & Sass, 2013). Most Proto-Canaanite signs look like either Phoenician or Cretan signs, but some notable exceptions exist. Some glyphs are hapax graphomena[1] in Phoenician texts but multiply attested in Crete or Cyprus. Simple graphemes such as straight or zigzag lines, circles, squares, triangles, and crosses, are not informative because they are attested in ancient writing systems worldwide.

Based on the hypothesis mentioned above, Proto-Canaanite is frequently used as a synonym of Proto-Sinaitic script[2] for specimens unearthed in Canaan, encompassing modern Lebanon, Israel, Palestine, and western parts of Syria. Proto-Sinaitic and Proto-Canaanite are viewed as subsequent steps of the evolution of linear writing to a phonetic alphabet. Some authors refer to early (Proto-) Sino-Canaanite scripts implying that specimens from the heart of Egypt record the same language as those from Israel and Lebanon. Synonymous are terms that mean the same thing or almost. To be synonymous, the terms in question must refer to the same script, i.e., comprise identical sets of characters for the same language. For example, we may claim that Modern English and American scripts are synonymous because they only differ in the dialectal use of a few letters. Classic Latin and Roman alphabets are synonymous because Romans used Latin. However, Latin, Swedish, and Modern English scripts are not synonymous.

Since the oldest securely termed alphabet, the so-called Phoenician, was found in Canaan starting from about 1050 BC, any writing found in the region before that date is considered proto-alphabetic. The series of equations: the alphabet is Phoenician; Phoenician is Canaanite; Canaanite is Sinaitic; therefore, Sinaitic is alphabetic is typically fallacious. Extending this logic, if a cross, a circle, and an unclosed rectangle found in Belgrade dated before 2000 BC, those also ought to be Proto-Sinaitic, ‘proving’ that Semitic speakers had already migrated to Serbia and invented their alphabet there. The date of the invention of the alphabet may be pushed back or forth, but its Semitic origin is never questioned.

The origin and linguistic membership of Sinaitic and Canaanite scripts will be eternally debated since phonetic transcriptions of ancient signs are not testable. Van Den Branden accepts Proto-Sinaitic as Semitic but explains a North-Arabian instead of a Canaanite origin (van den Branden, 1966). Others dispute the alphabetic nature of the Proto-Sinaitic script, arguing that the early inscriptions (see section The Proto-Sinaitic script) are proper Egyptian hieroglyphics (Morenz, 2011) or that no alphabetic inscriptions can be dated before the late14th-century BC with any confidence (Sass, 2005). As Sass remarks, one of the problems of an alphabetic interpretation of the Proto-Sinaitic inscriptions, or their early dating, is the long gap of place and time to the alphabetic boom observed after the 10th century. Sass considers this boom to have begun much earlier, around 1300 BC, based on four short inscriptions from Tel Lachish (today in south Israel) and a couple of other inscriptions from Israel dated to that period (Fig. 1). Since their discovery in the 1930s, the Tel Lachish inscriptions have been the missing link between the Proto-Sinaitic script and the first alphabet (Diringer, 1943). Nevertheless, Diringer warned that the evidence for an alphabetic nature of these scripts is weak, the gap is still enormous, and the provenance of genuinely linear scripts from Crete or elsewhere cannot be ruled out.

Figure 1. Proto-Canaanite inscriptions from Shechem (A), Tel Lachish (B, C, E, F), and Raddana (D), after Sass (1988), with the permission of Prof. Benjamin Sass. Similar or identical earlier symbols from Crete dated to 2100-1450 BC (Godart & Olivier, 1996; Salgarella & Castellan, 2021), framed in blue, and from Armenia, 7-6th millennium BC (Vahanyan & Vahanyan, 2009), in red, are shown in the margins.

Sass proposed his competition theory to justify the remaining 500-year lapse from the Proto-Sinaitic to Proto-Canaanite alphabet. The early alphabet would be suppressed by the influential writers and users of the long-established Egyptian and Akkadian systems in the Levant. The ban would have ended by the 13th-century allowing the alphabet to evolve from pictographs to linear graphemes (Sass, 2005). The problem with this theory is that it has no parallel in history. No abandoned alphabet has come to vogue 500 years later. Languages and scripts are quickly forgotten when not practiced. It would be the same writers who could use the new, convenient script.

Another inscription from Tel Lachish was found at an earlier stratum recently dated by radiocarbon measurements to the mid-15th century (Höflmayer et al., 2021; Webster et al., 2019). This finding somewhat shortens the gap between Proto-Sinaitic and Proto-Canaanite scripts. The inscription features two lines of three signs each. One of the signs is repeated as a separator between the lines. There are also two broken signs which may have conceivably been parts of a single sign. Again, all the signs closely resemble Cretan hieroglyphs or Linear A graphemes (Fig. 2).

Figure 2. Tel Lachish inscription dated around 1450 BC (Höflmayer et al., 2021). Artwork by J. Dye under Creative Commons license. Similar Cretan signs classified as Linear A (framed blue; SigLA) or hieroglyphics (red; CHIC) are shown for comparison: TY 2: A309^b, HT 31: AB58, KE Wc 2b: A702, HT 46b: A707, KH 9 AB11, CHIC #044.a, and CHIC #50a. The sign CHIC #113.b2 is an alternative candidate for the broken graphemes assumed connected.

According to the same study, the stratum continues through the first half of the 14th century BC, coinciding at least partially with the Amarna period. This period is known from a massive archive of diplomatic correspondence between the Egyptian administration and its representatives in Canaan, or neighboring leaders, written on clay tablets in cuneiform script around 1350 BC. The official script, if not the only one, in Canaan was Akkadian cuneiform by 1330 BC[3], while Linear A thrived in Crete from 1800 to 1450 and continued as Linear B until 1200 (Daniels & Bright, 1996). No cuneiform sign was ever found in Crete. Lachish was a substantial commercial and political center, predominantly mentioned in the Amarna letters and Egyptian papyri. Excavations revealed imports from Egypt, Cyprus, and the Aegean, attesting to the regional importance of the site (Höflmayer et al., 2021) but also suggesting the possibility of cultural relations and script influences from abroad.

By association with Egyptian hieroglyphs and on the alleged acrophonic principle of Proto-Sinaitic/Canaanite scripts, Höflmayer and colleagues interpret the first line as Hebrew ‘bd, meaning enslaved person, and the second as npt, honey, or tpn, to turn. Because slave-honey or slave-to-turn do not make much sense in the context of a sherd (probable fragment of a jar), they consider these sequences to be parts of personal names. They remark the existence of Hebrew names based on the root ‘slave’ but admit the second name to be unknown. Authors typically resort to personal names or theonyms when their reading makes no better sense. Conveniently, proper names do not need further explanation.

Instead, in Middle Egyptian (Dickson, 2006), tp (tip, top?) means the best of, and bd means natron, a mineral salt found in dried lake beds, consisting of hydrated sodium carbonate. The phrase the best-of-natron would cohere much better to a clay container of natron. Also known as washing soda, natron was a historical commodity in ancient Egypt[4]. It was harvested directly as a salt mixture from dry lake beds and used for thousands of years as a cleaning product for both the home and body. Blended with oil, it was an early form of soap. It softens the water and removes oil and grease. Undiluted natron was used as an antiseptic for teeth, wounds, and minor cuts. As natron dries and preserves fish and meat, it was also used for mummification in ancient Egypt. It was an ingredient of the distinct color called Egyptian blue and Egyptian faience, i.e., glazed and decorated earthenware (Fig. 3). Therefore, it is expected in pottery workshops, which eventually offer high pottery yields to archaeologists. In Roman times, it was also used for glassmaking.

Figure 3. A faience vase fabricated in part from natron, dating to the New Kingdom of Egypt (c. 1450–1350 BC). Artwork by the Walters Art Museum. Creative Commons license.

Fig. 2 discerns character-1 as a loop-like grapheme (e.g., Linear A TY2:A309b) or a cursive form of ‘ayin (Fig. 4) rather than a typically stylized closed circle (‘ayin). We may, therefore, use a wildcard for this character. Also, the broken graphemes 8 may be reconstructed as a pot-like pictogram (e.g., CHIC#113.b2, also attested in the Serabit inscription from Sinai; see section The Proto-Sinaitic script), meaning what it looks like, a pot. If Höflmayer and colleagues got the phonetics right, but the language of the inscription was Middle Egyptian (2000-1300 BC), we may read (Dickson, 2006):

<*BDNTPN*> container

<w>	*not*; property; district, region; they, them, their
<Abd>	month; monthly festival
<wbd>	burn, heat, be scalded
<bd>	*natron*
<n>	not; to, for, to (persons), in (sun, dew, time), because, belongs to; we, us, our
<nt>	of, belonging to
<t>	you, your; bread, loaf
<tp>	tip (top), the best of, upon, person, people (e.g., <tp Hsb> = correct method, norm, standard)
<tpn[fr]>	a good beginning
<pn>	this
<pn[a]>	turn upside down, turn
<pn[ana]>	turn over and over
<pn[ay]>	reversal?
<pn[q]>	bale out (a boat, water), swab up (blood), poke (at someone), expend (provisions)
<pn[s]>	clay

Combining the more sensible sememes (highlighted), we understand the inscription as the object’s specifications or user instructions:

W-BD = Semitic/Greek ‘BD = absence of natron, not-natron, natron-free, unglazed; N=for; T=bread; PN[S]= clay; sign-8=pot: non-glazed pot for bread (baking).

WBD= burn, heat, scald; N= not; TP= tip; N= not: tip/correct method for (not) burning/heating/scalding, do not…

WBD= burn, heat, scald; N=not; T= you; PN[*]= turn: to avoid burning, (you should) turn…

Such secular interpretations are preferred because, unlike proper names, these can be tested against the archaeological context or independent linguistic material.

There are only two glossed Middle Egyptian words fitting the [?bd] pattern. The Abd interpretation does not help in this context. We are left with the wbd solution, which suggests a w reading for sign 1, either alone (w bd; no natron) or as part of wbd (burn, heat, scald). The Egyptian grapheme <w> corresponds to Gardiner sign G43, which was pronounced as /w/, the voiced labial-velar approximant of English weep, or as a close back rounded vowel, /u/, as in the name of the builder of the Great Pyramid, Khwfw, (Khufu). In Coptic, which uses the Greek alphabet, w is written as ou and pronounced as /u/, probably because there is no /w/ in Greek and /u/ (οὐ; ou or oy; OY) was the closest alternative (Loprieno, 1995; Ross, 2022). Ou persisted throughout the history of Greek, from Homer to some archaic expressions still popular in Modern Greek. It has been recorded as the first name of Greek O, the name of Latin U, a diphthong, and 29 other uses, including the negative particle not (as in Middle Egyptian) and all other types of negation, denial, prohibition, negative reasoning, doubt, dispute, omission, negative question tags, double negations, etc. Therefore, we may predict that the diphthong may retain some of these negative meanings, negating what follows when found within words. The same applies, perhaps, to Latin U.

Figure 4. The archaic Cyrillic letter Uk is shown as a digraph (1) or a vertical ligature (2). A cursive form of Hebrew 'ayin (3) is also compared. Artwork by Пакко (digraph and monograph) under Creative Commons license, and Dan Pelleg ('ayin) in the public domain.

Before vowels, ou becomes ouk (οὐκ; oyk; /uk/), keeping all its negative meanings. The Greek ligature Ou (Ȣ ȣ), resembling the Linear A sign A309b, is frequently encountered in medieval Greek manuscripts and in modern editions of classical texts. This ligature is still used in cursive Modern Greek, though not as frequently as before. However, the digraph Oy and the ligature (Fig. 4) were borrowed by the early Cyrillic alphabet directly from Greek, in the 9th century AD, and named Uk. But it was replaced by the Cyrillic Y in Old Russian by the 15th century AD.

Therefore, an ou reading of Linear-A sign A309b is plausible. I have also argued that the Greek rough breathing diacritic (spiritus asper) frequently denotes something missing, elision, omission, absence. Most Greek words starting with Y take this diacritic, probably to suggest the omission of O from an original Oy. It may correspond to Semitic ‘ayin, which is transliterated as an apostrophe and pronounced as a soundless glottal stop (missing sound). It all looks like the Greek negation ou derived directly from Egyptian w, but its graphic transcription first appeared in Crete.

References.

Daniels, P. T., & Bright, W. (1996). The World’s Writing Systems. Oxford University Press.

Dickson, P. (2006). Dictionary of Middle Egyptian. Open-source.

Diringer, D. (1943). The Palestinian Inscriptions and the Origin of the Alphabet. Journal of the American Oriental Society, 63(1), 24.

Finkelstein, I., & Sass, B. (2013). The West Semitic Alphabetic Inscriptions, Late Bronze II to Iron IIA: Archeological Context, Distribution and Chronology. Hebrew Bible and Ancient Israel, 2, 149–220.

Gardiner, A. H. (1916). The Egyptian Origin of the Semitic Alphabet. The Journal of Egyptian Archaeology, 3(1), 1-16 (23 pages).

Godart, L., & Olivier, J.-P. (1996). Corpus Hieroglyphicarum Inscriptionum Cretae. In Études crétoises (Vol. 31). École française d’Athènes.

Höflmayer, F., Misgav, H., Webster, L., & Streit, K. (2021). Early alphabetic writing in the ancient Near East: the ‘missing link’ from Tel Lachish. Antiquity, 95(381), 705–719.

Loprieno, Antonio. (1995). Ancient Egyptian : a linguistic introduction. Cambridge University Press.

Morenz, L. D. (2011). Die Genese der Alphabetschrift: ein Markstein ägyptisch-kanaanäischer Kulturkontakte. Ergon-Verlag.

Ross, K. L. (2022). The Pronunciation of Ancient Egyptian. The Proceedings of the Friesian School.

Salgarella, E., & Castellan, S. (2021). SigLA: The Signs of Linear A. A Palæographical Database. Grapholinguistics in the 21st Century, Part II, 5, 945–962.