Proto-Canaanite scripts

Proto-Canaanite refers to Canaan’s most ancient linear scripts, i.e., modern-day Israel and Lebanon. These scripts are essential to link a couple of short allegedly alphabetic inscriptions from Sinai and south Egypt (dated to 1900-1500 BC; see section The Proto-Sinaitic script) and the full-blown Phoenician alphabet conventionally dated after 1050 BC. By 1330 BC. The official script in Canaan was Akkadian cuneiform, as manifested by the massive Amarna archive of official correspondence[1]. Meanwhile, Linear A thrived in Crete from 1800 to 1450 BC, preceded by substantially linear Cretan hieroglyphs (2100-1700 BC). Linear A continued as Linear B until 1200 BC (Daniels & Bright, 1996). The oldest Proto-Canaanite inscription, radiocarbon-dated to 1350 BC, was found in Tel Lachish (see section The Tel Lachish script). The Proto-Canaanite corpus consists of some 20 extant inscriptions or fragments totaling no more than 1300 characters, 1214 of which are classified as pseudo-hieroglyphs of the Byblos syllabary (Garfinkel et al., 2015; Vita & Zamora, 2018). For comparison, about 6400 documents with more than 65000 linear signs were found in Crete, mainland Greece, and the Aegean islands, dating from the late 3rd millennium to 1200 BC (Petrolito et al., 2015).

The so-called Byblos pseudo-hieroglyphs (later, Byblos script) constitute the core of the linear Proto-Canaanite writing system. In the late 1920s to early 30s, the French archaeologist Maurice Dunand discovered in Byblos, Lebanon, 16 texts of length ranging from 7 to 68 characters (Dunand, 1945; 1978). Two texts were hammered on metal tablets, five on metal spatulas, and nine on limestone slabs. The longest inscriptions are shown in Fig. 1.

Figure 1. Tablet inscriptions in Byblos pseudo-hieroglyphic script, after Dunand (1945). A. The first side of the inscription, c. B. The two sides of the inscription d.

Dunand recognized 123 distinct signs (Fig. 2) in the newly discovered script. This number was more extensive than needed for a simple syllabary. He, therefore, classified the script as logo-syllabary, i.e., a writing system using syllabograms and logograms, analogous to Egyptian hieroglyphic and Linear A systems. In a logo syllabary, signs are used as syllables and carry meaning. He originally dated the inscriptions to the beginning of the 2nd millennium BC, based on his intuition and the similarity of some characters with the Egyptian hieroglyphs. At the same time, he thought the oldest typical Phoenician inscription on the Akhiram sarcophagus dated to about 1300 BC. Because of the geographical coincidence and the chronological succession, Dunand concluded that the Byblos script was the origin of the Phoenician alphabet. However, one of the items he found was inscribed with pseudo-hieroglyphs and classical Phoenician letters, indicating that both scripts were used in parallel for some time. Sass criticized Dunand’s hypotheses and down-dated the artifacts to 900-830 BC and the Phoenician alphabet to 850 BC (Sass, 2019).

Figure 2. A. List of Byblos pseudo-hieroglyphics according to Dunand (1945). Corresponding Cretan (Minoan) signs and Old Balkan signs (framed in red) not identified by Dunand are juxtaposed. The blue asterisks indicate signs for which Dunand identified Cretan counterparts. B. Cretan counterparts of Byblos pseudo-hieroglyphs after Dunand (1945).

Something strange is going on with the Byblos script. We can identify languages by their pattern of letter usage. The frequency at which letters are used has not changed from Ancient Greek to Modern Greek, with few exceptions due to known recent reforms (Fig. 3). For example, the adoption of Demotic vocabulary and grammar in the 20th century caused feminine nouns previously ending in -ις (-is) to end with -η (-ē). This reform caused the frequency of usage to increase for H and decrease for I and S. The patterns of letter usage in Greek, Latin, French, and English are conspicuously different. The usage frequency patterns for single letters and small clusters (digraphs, trigraphs, syllables, or other morphemes) may be precisely determined from dictionaries and used to compute graphocentric ‘phylogenies’ of languages.

Figure 3. Standardized frequencies of common letters in Ancient Greek (AG), Modern Greek (GR), Latin (LA), English (EN), and French (FR).

This type of data and method should assist in testing hypotheses about the language behind an undeciphered script. For example, if Linear B is truly Greek, the usage frequency of its signs should significantly match the frequencies of usage of the corresponding syllables in Homeric Greek, at least. The Appendix describes a principal components analysis of the frequencies of letters, bigrams, and trigrams in lexical corpora of Ancient Greek, Modern Greek, Latin, French, English, The Iliad, The Odyssey, Theogony, and Works and Days. According to this analysis, 93.3% of the letter and cluster frequency variance is non-specific (principal component 1; PC1). The second dimension (PC2) separates the Greek lexica from the Latin-based ones, accounting for 3.6% of the variance. The third dimension (1.9%) separates Ancient and Modern Greek from the epic poetry vocabulary and the fourth (0.6%) Modern English and French from Latin. The fifth dimension (0.3%) separates English from French, and the sixth (0.2%), Modern Greek from Ancient Greek. The seventh dimension (0.06%) separates Homer (The Odyssey and The Iliad) from Hesiod (Theogony and Works and Days). The eighth dimension (0.005%) separates Theogony from Works and Days.

The Odyssey and Iliad remain inseparable. If any, their differences in letter usage represent less than 0.005% of the overall variance. The results also suggest that Modern Greek is closer to Ancient Greek than written English is to written French. Ancient Greek vocabulary is closer to Modern Green than epic poetry. Works attributed to Hesiod differ between them more from those attributed to Homer.

If the Byblos script records a Semitic language, the frequencies of its signs should correspond to the frequencies of the syllables of Semitic languages. The most frequent Byblos sign would correspond to the most frequent Semitic syllable, the second to the second, etc. Of course, attributing the so-identified corresponding syllables to the signs of the ancient syllabary, the text must begin making sense.

Figure 4. Frequencies of Byblos-script signs in Dunand’s (1946) first 10 published inscriptions. Data by Hans van Deukeren (2014). A. The most frequent signs. B. Signs occurring more than once but less than 15 times.

Hans van Deukeren (2014) counted the occurrences of each Byblos sign in the first ten inscriptions published by Dunand in 1946 (Fig. 4). The frequencies of the 32 most frequent signs drop smoothly from 63 (0.58%) to 15 (0.14%). The frequency of the following sign drops abruptly to 11 (0.10%). There are 89 signs seen at least twice and another 40 signs seen only once. The latter contain unclear or broken signs, perhaps some variants of more frequent signs, special characters, and genuinely rare graphemes. We may hypothesize that the 32 most frequent signs correspond to the most frequent Semitic phonemes or combinations. Comparing the two Byblos tablets, both with statistically sound numbers of characters, we notice that some letters are pretty regular in one tablet but absent from the other. For instance, the character G7 (Fig. 2) appears seven times in the specimen shown in Fig. 1A but is missing from the model in Fig. 1B. If the two tablets record the same language, such variation may be explained as scribe specific. Scribes appear to choose which graphemes to use or how to write them. The script seems to have been in perpetual experimental reform.

Dunand had noticed striking similarities between the pseudo-hieroglyphic, on the one hand, and Egyptian hieroglyphs or Cretan Linear A signs, on the other. His comparative table (Fig. 2B) shows one-third of the Byblos signs corresponding to older signs from Crete. Today, this correspondence is raised to about 50% of the Byblos script (Fig. 2A). Nine more signs are identical to Neolithic glyphs found in Balkan sites dated millennia earlier (Winn, 1981). With many of the remaining signs identifiable as contemporary, if not earlier, Cypro-Minoan (Everson, 2020) or typical Egyptian hieroglyphs, few are left to consider Byblian innovations. Dunand explained that people might invent simple characters independently. Many people may draw circles or other simple geometrical signs without necessarily imitating each other. He, therefore, favored the Egyptian connection.

However, there are infinite possibilities for a sign with two or more strokes. The probability that two people will draw the same pattern without looking at each other is infinitely small. One should also contemplate the likelihood that two independent character sets will present an overlap of 50%. When drawing complex pictograms of animals, human parts, pots, or other objects, the results will be similar as far as the thing is recognized. Figurative pictograms may be considered independent inventions, and the overlap between the symbolic Byblos signs and the Egyptian hieroglyphs may be attributed to chance.

A different Proto-Canaanite script was discovered in Khirbet Qeiyafa, a small, fortified, early Iron Age site on the geographic frontier between Iron Age Judah and Philistia, Israel, and the chronological transition from Israeli Iron Age I (1200–1000 BC) to Iron Age IIA (1000–925 BC), i.e., around 1000 BC (Garfinkel et al., 2018; Levin, 2018). Today, this script is represented by three clay inscriptions of significant length found in the same archaeological stratum (Fig. 5). The archaeological context was radiometrically dated to 1020-980 or 1018-948 BC (Garfinkel et al., 2015). One of the shortest inscriptions, known as the ’Išbaʿal inscription, was incised on a jar before firing, presumably by a professional potter. The longest was written with ink on an ostracon. The ʿIzbet Ṣarṭah ostracon is, perhaps, the most interesting because it bears a version of the alphabet used.

Figure 5. Inscriptions from the same stratigraphic layer of Khirbet Qeiyafa, Israel, dated 1020-980 BC (Garfinkel et al., 2015). A. The ’Išbaʿal inscription (Garfinkel et al. 2015; drawing by A. Yardeni). B. Khirbet Qeiyafa ostracon (Misgav, Garfinkel, and Ganor 2009). C. ʿIzbet Ṣarṭah ostracon (Sass, 1988). With the permission of Prof. Benjamin Sass.

The letters had not yet gained the consistent stance characteristic of an alphabet. We note the presence of typical Linear A/B signs or Old European signs, which are absent from the Byblos pseudo-hieroglyphic script but present in latter standard Phoenician inscriptions. The sign 16 of the second line of the ʿIzbet Ṣarṭah ostracon (𐤈; ṭēt; Theta) also appears ninth of the bottom line of the ʿIzbet Ṣarṭah ostracon and twice in the second line of the Khirbet Qeiyafa ostracon. It is identical to the Minoan/Mycenean sign AB77 (𐌈). It dates from the earliest period of Vinča culture (w), 6th-5th millennia BC (Ager & Paliga, 2021), and has also been found among petroglyphs of the 7th-6th millennia BC in Armenia (Vahanyan & Vahanyan, 2009). The five-step ladder sign at position 23 of line 4 of the Khirbet Qeiyafa ostracon is the AB57. The perfect circles with or without central points are frequently used in the Khirbet Qeiyafa ostraca and Byblos pseudo-hieroglyphic inscriptions but not in proto-Sinaitic or Tel Lachish specimens. They correspond to Cretan sign A309a (𐤏) and its variant A309c (𐰧), existing already in Crete since the LMIB period (1500–1450 BC; Fig. 2B) and, of course, in the Balkans much earlier (DS167 and DS171). The sign that became het (𐩳; e.g., ʿIzbet Ṣarṭah line 4 sign 6) existed in Crete as Linear A360 since MMIIB (1750-1700 BC), but it was also prevalent in various figurative hieroglyphic forms as CHIC#038 (2100-1800 BC) or as DS65 (6th-5th millennia BC) in the Balkans. The Σ-like sign A305 takes its final simplified form as used later in most archaic and Classical Greek alphabets. But this form was not adopted in the Classical Phoenician script. The same applies to 𐊵 (ʿIzbet Ṣarṭah line 2 sign 1; Cretan AB27). The 𐌓 (ʿIzbet Ṣarṭah line 4 position 25) was used in Crete as A358 since MMIIB (1750-1700 BC). There are also several Linear A signs A349 on the Khirbet Qeiyafa ostracon.

The ʿIzbet Ṣarṭah inscription was dated by letter typology to the Early Iron Age IIA, i.e., 10th century BC, with a question mark; perhaps a little earlier (Finkelstein & Sass, 2013). A particular interest in the inscription lies in its bottom line, an abecedary with the letters written almost in their later conventional order. The Y has moved after O, the letters Q (𐌒), P (𐤐), and R (𐩾) were replaced with two identical q-like signs (𐩺), M and N are swapped and hardly recognizable, while there is a ᛚ sign in the place of 𐊑 (Ksi or Ṡāmek). An S-like form was restored as a rounded 𐩦 (Phoenician Sīn rotated 90° anti-clockwise).

Sass and colleagues (Sass et al., 2015) characterize the ‘Izbet Ṣarṭah inscription as ‘ineptly incised’ and cite Joseph Naveh (Naveh, 1978) commenting:

The ʿIzbet Ṣarṭah ostracon contains the attempt of an unskilled person … to write an abecedary with the twenty-two Canaanite letters.... His confusion of letters and his mistakes seem to be so serious that I would not recommend the drawing of palaeographic conclusions from any of the forms produced by him. We cannot know which letter forms are based on the contemporary scribal tradition and which are the products of either the writer’s poor training or his bad memory.

The sherd of pottery (ostracon) was inscribed after firing, which means that the scribe may have been a layperson – not the potter – a trainee, student, or school child learning the alphabet. Other inscriptions from the same area and time seem to have been written with chalk (Finkelstein & Sass, 2013). Writing is an art, and writers have their own artistic style. In an era when writing is supposed to be scarcely scattered and in the invention process, writers cannot be expected to use standard methods. The ʿIzbet Ṣarṭah scribe cannot be accused of bad memory when we do not know how often she had the chance to see other people’s writings or for incompetence in predicting how the alphabet will eventually be fixed. The ostracon reminds us that handwriting is highly variable, like everything else we do by hand. It may contain ‘errors,’ unique features, and innovation. The ‘Izbet Ṣarṭah inscription shows the state of development of writing in Canaan in the 10th century BC.

Figure 6. The Ahiram graffito dated to 13th century BC (A) with Cretan Linear A transliteration (B) from 1800-1450 BC and the corresponding Danube-script signs from 6000-5000 BC (C). The Linear A signs are: A707 (1); A353 (2); A309a (3); AB02 (4); CHIC#40.b;  AB27 (6); AB47 (7); A706 (8); A349 (9); AB302 (10); AB56 (11); and A319 (12). The Proto-Canaanite signs (A) 1, 2, 5, 6, 8, and 11 are identical to Danube-script (C) signs DS138, DS51, DS87, DS2, DS237, and DS65, respectively; signs 3 and 4 are commonly found in all archaic scripts (Godart & Olivier, 1996; Martirosyan, n.d.; Salgarella & Castellan, 2021; Vahanyan & Vahanyan, 2009; and Winn, 1981). Artwork by Pierre Montet and Abdnr. Marked as public domain. 

Figure 7. Inscription of the Ahiram’s sarcophagus, Byblos, Lebanon, 10th century BC. Artwork by anonymous. Marked as public domain.

The graffito of Fig. 6 is yet another example of Proto-Canaanite writing. It was found halfway down the shaft of the tomb of Ahiram in Byblos, Lebanon, housing the homonymous sarcophagus with the oldest known typical Phoenician inscription (Fig. 7). The French archaeologist Pierre Montet (1885 – 1966) excavated the royal tomb between 1921 and 1924, sketched the inscription, and photographed it. The text has been interpreted to warn against looting and dates to the 13th century BC. Unfortunately, the sketch and the photograph present discrepancies at the level of the crucial sign 7, i.e., the fourth letter of the middle line reading from the right and the second letter of the bottom line (sign 11; Dussaud, 1924). The more recent version of the sketch shown here (Fig. 6A), currently in the public domain, is still different. It is explained that the fourth letter of the middle line represents two overlapping letters, a bēt (𐤁) and a qoph (𐤒)[2]. I do not know which version is closer to reality. The earlier graphical interpretations assumed that the inscription was Phoenician and that the Phoenician alphabet did not have that very letter. If the current version is more accurate, I would say sign 7 is the Linear A and B sign AB47[3], and it is complex enough to leave no doubt about its identity. The second glyph of the middle line from the left (sign 6) has been found in many Aegean and Anatolian inscriptions and is a variant of Linear A sign AB27 (𐤶). It is identical to Danube-script sign DS2. It also appears in the inscription of Ahiram’s sarcophagus (Fig. 7), where it is usually interpreted as K. K is read ‘aleph, while the classical Phoenician’ aleph does not exist in those inscriptions. The other signs in the Ahiram’s graffito are also recognized as Cretan Linear A and Danube-script signs. Therefore, Ahiram’s inscriptions resemble more Danube script or Linear A, because of sign 7, than the Classical Phoenician script.

Curiously, the anagrams ʿbd/dbʿ and bdʿ/ʿdb – with the sign ʿ rendering the Semitic ʿayin (Phoenician O) – appear on a Tel Lachish clay (see section The Tell Lachish script), the ’Išbaʿal inscription, and at least twice on the ʿIzbet Ṣarṭah ostracon, which also features the bʿd/dʿb cluster. Yosef Garfinkel and colleagues who dated and described the ’Išbaʿal inscription, interpreted it as’ Išbaʿal son of Bedaʿ (read from right to left), hence the name of the specimen (Garfinkel et al., 2015). The authors concentrate longer on’ Išbaʿal and find it equivalent to Biblical Eshbaʿal, a rival of King David, who is supposed to have lived around 1000 BC. They explain that names built on the theophoric root baʿal must have been common in David’s times but disappeared from the Bible and Judah inscriptions later. Since the jar dates around 1000 BC, their’ Išbaʿal reading is consistent with the Bible and does not raise any contradiction with Judaean onomastic tradition. We must understand that the Bible justifies their reading, and the inscription confirms the Bible. Since the jar dates around 1000 BC, their’ Išbaʿal reading is consistent with the Bible and does not raise any contradiction with Judaean onomastic tradition. (Golub, 2014). The name Bedaʿ, they say, is unique. In other words, the authors read something that has never been seen in Hebrew before or ever after. Yet, they are confident that the inscription is in Biblical Hebrew because a three-letter root appears once or twice in the Bible. They finally interpret the inscription as a trademark; the jar’s content comes from the town/estate of’ Išbaʿal Bedaʿson. Trademarks are immune to criticism or critique.

Nevertheless, Garfinkel et al. (2015) recognize that bdʿ means producing, inventing, or beginning in Arabic. The cluster bd alone is a contraction of byd, meaning from, or by the hand of (Benz, 1972). Considering the ‘ayin (‘) as an ellipse, something missing, or an abbreviation, the authors suggest an alternative Arabic interpretation of bd ‘, ‘[the deity] has created.’ The [deity] insert is, of course, not necessary. The creators of the jar certainly knew very well who had created it. Marks like ‘made by’ or ‘made in’ are seen in today’s products. As I have mentioned (see section The Tel Lachish script), the Egyptian equivalent of ‘ayin, <w>, means not (compare Greek οὐ; ou or oy; OY; /u/; negative particle), or property, region, district (compare Greek οὗ; ou; where; and French où; /u/ where). The compound bd-‘would, thus, mean the place where something originates or was created. The toponym remains to be identified.

Figure 8. The ’Išbaʿal inscription flipped and rotated 180°.

Could the ’Išbaʿal inscription be Greek? If we face the orifice of the jar and turn it 180°, we obtain the lower part of Fig. 8. We may, thus, identify the characters AΣbO)I[Σ?]IbDO. Taking character 5 as an L (as do Garfinkel et al.), this sequence transliterates as ASBOLI[S?]IBDO. A search for Ancient Greek words containing ASBO or IBDO gave plausible results. The word ἀσβόλη (asbolē; ASBOL-H) and the more Attic form ἄσβολος (asbolos; ASBOL-OS) mean soot. The verb ἀσβολόω (asboloō; ASBOL-OΩ) mans to cover with soot. The word ἀσβολθέν is glossed by Hesychius as μέλαν (melan), black, dark, swarthy, black marks, murky, obscure, hence Modern Greek μελάνι (melani), ink. Brief, all the words containing the root asbol are semantically related to soot. Ancient inks were made of soot (Tsuen-Hsuin, 1985).

Among the few words containing -ibdo- we find κίβδος (kibdos; KIBDOS), meaning alloy or dross, i.e., waste, impure matter, waste product from working with metal, worthless or trivial matter, junk, rubbish. Therefore, ASBOL*IBDO* could mean ashtray, ash-bin, ash-trash, dustbin, a covered container for collecting soot, ashes, or burned waste. However, the broken grapheme 8 looks more like Σ than K. The word στιβδός (stibdos or ςibdos; StIBDOS; Hesychius στικτός, stiktos) means pricked, tattooed, spotted, dappled, i.e., marked with spots or round patches. In this case, ASBOL-SIBDO would mean marked, stained, decorated, with black, soot ink. The letter Stigma (ς) is supposedly a ligature of S and T. There is evidence that Stigma and Sigma have been confused in history and may have sometimes been represented with the same grapheme. For example, the Greek terminal S has the shape of Stigma (ς) and has probably functioned as a word endpoint (dot, stigma). Their phonetic difference may have been dialectal. Both instances of Σ in ASBOL-SIBDO may have represented Stigma pronounced /ʃ/ (as opposed to Ancient Greek San, /s/) like in English ash or Semitic’ Išbaʿal. The compound words ἀκρομόλιβδος (akro-molibdos), meaning with lead at the edge, and κυκλομόλιβδος (kyklo-molibdos), round lead-pencil, suggest that the stem ibdo may well signify a writing or drawing tool. A writing or drawing tool that works with shoot ink would be ASBOL-SIBDO.

The Greek interpretation produces two coherent words with repeatedly attested roots. It flows freely from the form of the intact letters without the need to distort their phonetics or add arbitrary vowels among them. This hypothesis generates testable, in principle, predictions about the jar’s content or decoration. It does not resort to unfalsifiable proper names and theistic logic. It would imply that other Proto-Canaanite inscriptions may also be Archaic Greek. All the letters of the ’Išbaʿal inscription and the other contemporary Canaanite inscriptions exist in Archaic Greek epigraphy (Salgarella & Castellan, 2021). Although nobody seems to deny this, the direction of cultural exchange is generally agreed from Canaan to Greece, based on Herodotus’ misunderstood text (Hdt. 5.58). With the re-interpretation of Herodotus’ term ‘Phoenicians’ and the related myths (see chapter Phoenician myths), the transfer of writing from an ethnic or geographical Phoenicia remains without literary support. When Greek texts are read with Semitic phonetics in the wrong direction, they make no secular sense. Theistic and onomastic interpretations are the only resort.

The most sensible interpretation of the ’Išbaʿal inscription is obtained, however, assuming a phonetic transcription from Ancient Egyptian with 𐰧 (O) interpreted as /w/ and sign 7 restored as 𑀏 (D) rather than 𐋉 (B). From left to right, the inscription would so read ODBISDILOBSA. Middle Egyptian dbi means stop up, sd means clothe, and bsA means protect. The phrase would mean something like ‘cover (plug) with a cloth to protect’ (the content). In Late Egyptian, there is also sdi, chaudron.

According to phonocentric theory, Greek is basically an Indo-European language installed in Helladic territories with the arrival of the Greeks from somewhere else. Most scholars date the coming of Proto-Greek to the transition from Early Helladic II to Early Helladic III, around 2400−2200/2100 BC (Meier-Brügger, 2017), i.e., before the appearance of any linear scripts in the area. The diversification into northern and southern groups would have begun by approximately 1700 BC. But all the scenarios are open. Coleman argues for a first wave of Indo-European Pre-Greek speakers migrated to Greece around 4500-4400 BC. Therefore, the Pre-Greek substrate was Indo-European. Then, a second wave of Proto-Greeks arrived around 3200 BC (Coleman, 2000). That would be consistent with the Kurgan hypothesis of successive migrations of Pontic steppe peoples towards the West, based on cultural finds (Gimbutas, 1986, 1997). The Kurgan hypothesis would explain the resemblance of Danube, Vinča, Balkan, and North Hellenic signs to yet older Armenian petroglyphs (Vahanyan & Vahanyan, 2009), on the one hand, and to later Minoan and Mycenean linear scripts, on the other (see section Cretan scripts). Around the 12^th century BC, Aegean populations, probably from Crete, would have settled in Philistia and the Levant carrying with them their art and their writing system (Augustyn, 2020).

Figure 9. Ripple in water is a surface wave. Artwork by Agustín Ruiz. Creative Commons license.

Alternatively, there have been no massive migrations. Like all cultural systems, languages do not need migration to expand but are transited horizontally by social contact (Drinka, 2013; François, 2015). Wave models (Fig. 9) are more appropriate than phylogenetic trees to describe linguistic phenomena. Linguistic innovations act as secondary areal wave sources causing reflection or diffraction of the initial expansion like openings or obstacles do to mechanical waves according to the Huygens–Fresnel principle (Fig. 10). Because linguistic innovations occur at the microscopic level of linguistic units (phonemes, graphemes, lexemes, sememes, etc.) languages do not travel in space and time as solid entities but are subject to interference from multiple centers of innovation. We should not talk about the origin of a language but the origin of individual phonemes, graphemes, lexemes, sememes, and so on. Drinka’s mixed wave and tree models support Renfrew’s proposition that Greek – and other Indo-European languages – developed from an autochthone or Anatolian language, probably Luwian, via Minoan and Mycenean (Renfrew, 1973).

Figure 10. Wave refraction following the Huygens–Fresnel principle. Artwork by Arne Nordmann. Creative Commons license.

Indeed, if Indo-European languages began around the Black Sea, say in Armenia as the writing record would suggest, they would propagate towards North, South, East, and West in a wave fashion. We should, therefore, expect similarities between the oldest Armenian petroglyphs from the center, and the Egyptian hieroglyphs (Martirosyan, n.d.), Mesopotamian proto-cuneiform scripts (see section The first writings), Danube and Balkan scripts (Vahanyan & Vahanyan, 2009), early Chinese scripts (Fig. 11), and Cretan, Cypriot, and Levantine hieroglyphic or linear scripts, from the periphery. The phonemes, or the more directly attested graphemes from each cardinal point, are subject to independent innovations. Each innovation becomes a new center of linguistic radiation giving rise to discrete languages and writing systems, eventually alphabets. Each linguistic entity becomes a new wave source and continues to develop, expand, defuse, and diffract, subject to interference from neighboring centers.

Figure 11. Shang dynasty Chinese inscription on an oracle bone circa 1200 BC. Artwork by anonymous. Creative Commons license.

In all linguistic phylogenetic trees I have seen, Greek’s nearest neighbor is Armenian. Transposing this information to a wave-like linguistic expansion model with Armenian in the center, the first secondary Indo-European center of radiation would be Greek. Unfortunately, non-Indo-European languages are not usually included in Indo-European analyses. In any case, Greek is a mixture of an Indo-European language with a Pre-Greek substrate, the origin of which is debated. Beekes (Beekes, 2014; Beekes & Beek, 2010) identified 1106 Greek words of which Indo-European reconstruction seems impossible and labeled them as Pre-Greek. Among them, we find extremely common terms such as theonyms (Apollo, Artemis, Dionysus, Atlas, Athena, Asclepius, etc.), mythological characters (Achilles, Odysseus, Cadmus, etc.), toponyms (Attica, Crete, Olympus, etc.), plant names (e.g., the words for olive, vine or grape), as well as anatomical (neck, jaw), architectural (clay, stone, brick, labyrinth), geographical (mountain top, river bank, gorge, cavern), technical (copper, iron, gold, lead, spindle, cord, mesh, pickaxe, bridle, rein), biological (animal names), maritime (sea, to stir, govern), domestic (olive oil, oil flask), social (king, ruler, contest, sporting activity), musical (lyra, string or wind instruments), and weaponry terms (corselet, whip, javelin).

An implicit explanation is that local Pre-Greeks were familiar with all those things and had words for them. The Indo-European Greeks arriving in waves from the Russian steppes had not seen vines, olives, or the sea before. They were not aware of local Helladic heroes, gods, islands, and places. They were, therefore, not expected to have had words for them and had to borrow the local terms. However, not everything can be explained this way. Elementary history books tell us that iron was used only after 1000 BC. Why would Pre-Greeks have a word for iron before the arrival of the Greeks 1500 if not 3500 years earlier? Do we have words for things that will be discovered or invented by non-English speakers in 1500 years from now? Would an American scientist look up in long extinct indigenous American vocabularies to find a name for a material, tool, or method, she just developed? For such terms, Pre-Greek must be understood as contemporary to Greek but foreign. If Pre-Greek, iron was discovered and named outside the Hellenic world. Greek borrowed the term in historical times.

Beekes observes that certain consonant clusters, /bd/ and /sb/ among them, do not exist or are very rare in Indo-European languages but common in Pre-Greek. He does not specify where they come from. Plausible origins are Ancient Egyptian and the Semitic languages of the Levant. The cluster bd appears several times in the few Proto-Canaanite inscriptions presented above. From the so-called’ Išbaʿal inscription, I deduce that ASBOL means ash or soot and [S]IBDO means decorated, covered, coated. In Middle Egyptian (Dickson, 2006), we find the words <Asb> (/ɑ(ː)sb/) meaning glowing (of radiance), <Asbyw> (/ɑ(ː)sbiːuː/), flame, <psi> (/psi(ː)/), to cook and <wbd> (/w|uːbd/), to burn, heat, be scalded. In Late Egyptian (Lesko, 2002), flame becomes <bsw> (note the SB/BS inversion), <ps> means to bake, to cook, baker, boiler (of oil), roaster, loaf, <wbd> retains the meaning to burn, roast, <Sibd> means stock, and <Ibdw> means fish.

In Greek, the double phonemes /ps/, /bs/, /vs/, and /fs/ are represented by the grapheme Ψ. These are found in Ancient Greek ἕψω (‘epsō or hepsō), to boil, cook, digest, smelt, ἑψητός (‘epsētos or hepsētos; /epsi(:)tos/), boiled, boiled fish, or Modern Greek ψητός (psētos; /psi(:)tos/), roast, baked. The ending morpheme -εψω (-epsō) forms the future tense of hundreds of Greek verbs ending in -ευω (-eyō; /evo/), which reminds the also very common Egyptian ending -yw[1]. The inversion bs/sp, or ps/sp, holds in Greek for lighting up and lighting down/out. The notion of quenching is in Greek ἑσπέρα (‘espera or hespera) and its derivatives meaning evening, nightfall, night, hence, west. Whereas eps is a future morpheme, the inverse esp forms aorists. Compare, for example, the Homeric future ἕψομαι (‘epsomai or hepsomai) with second aorist ἑσπόμην (‘espomēn or hespomēn) from ἕπω (‘epō or hepō), to follow, accompany, escort, attend, pursue, follow up, especially in mind, understand, succeed, etc. Yet another example of antonymy by the ps/sp inversion involves the Ancient Greek ἁψίς (‘apsis, hapsis), loop, mesh, net, felloe of a wheel, the wheel itself, dowel-pin, arch or vault, triumphal arch, orbit, and ἀσπίς (aspis), shield. Whereas hapsis implies passing through it, aspis forbids it; it repels objects falling on it.

If ASBOL was Egyptian and ASB meant flame (fire), we should look for possible Egyptian meanings of OL, probably transliterated as <wl>. I found the Late Egyptian word <wl> meaning condition (Lesko, 2002). The stem <bwl> also exists in Late Egyptian with relevant meaning, to make firm, magnificence, from the verb to magnify, i.e., make big, make look bigger than it really is, expand. But Late Egyptian developed after 1350 BC (the Amarna Period) and may well have had Greek borrowings. Then, *<asbwl> would mean something like flame-conditioned (fired), fire-strengthened (made firm), or fireproof. The trouble is that bol also exists in Greek as in βόλος (bolos), glossed by Hesychius as clay, or βῶλος (bōlos), lump, a clod of earth, or clay. Then, a Greek asbol would mean fired clay; if asb- comes from the Egyptian <Asb> and carries the original meaning of glowing flame. The stems bwl, bol, and bōl, also remind the English bowl, Old English, Old Norse Bolle, and Old High German Bolla, all meaning pot, cup, bowl, which is typically made of fired clay. The Online Etymology Dictionary suggests that bowl and its old European cognates derive from Proto-Germanic *bul- (a round vessel), ultimately from PIE *bhel- (2), to blow, swell.

The PIE root *bhel- (2) refers to various round objects and to the notion of tumescent masculinity. Its reconstruction was based on a collection of morphologically, phonetically and semantically disparate words including English balloon, ballot, bawd, bold, boll, bollocks, bollix, boulder, boulevard, bull (bovine male animal), bullock, bulwark, follicle, folly, fool, foosball, full (to tread or beat cloth to cleanse or thicken it), pall-mall, phallus, Greek phyllon (leaf), phallos (swollen penis), Latin flos (flower), florere (to blossom, flourish), folium (leaf), Old Prussian balsinis (cushion), Old Norse belgr (bag, bellows), Old Irish bolgaim (I swell), blath (blossom, flower), bolach (pimple), bolg (bag), Breton bolc’h (flax pod), Serbian buljiti (to stare, be bug-eyed), and Serbo-Croatian blazina (pillow). The semantic relation of a clay container (bowl) with the bull, the flower, the phallus, the pillow, and the bellows may be questioned. The proponents do not explain how the specifications of *bhel- into blow and bowl came about. Why and how does the metathesis of L produce such profound semantic change? The more words we consider for the reconstruction of a root and try to explain, the more complex, vague, and dubious the phonocentric etymologies become. Instead, the above ‘out of Egypt’ hypothesis is based on archaeological evidence, historical considerations, and only evident assumptions about the phonetic evolution of the Egyptian <w>. The following graphocentric hypothesis would explain the semantic effects of the metathesis of L from BOWL to BLOW and suggest that this metathesis could have easily been elaborated in Europe using Archaic Greek (or “Phoenician”) graphemes.

Figure 12. Archaic Greek versions of Gamma, Lambda, and Ypsilon from Salgarella and Castellan (2021)

Fig. 12 compiles the graphemes representing the letters Gamma, Lambda, and Ypsilon as attested in Archaic Greek epigraphy. It shows that these letters had forms of angular or curved brackets and could be used as arrows to indicate direction. A curved, parenthesis-like form of L is used in the ’Išbaʿal inscription (grapheme 5), while Y was curved as U in Latin and as u in lowercase Greek. Curved forms are easier to produce in cursive writing, whereas angular forms are easier to carve. We also note, from Fig. 12, that these three letters could be confused. Modern γ could be read as Y; Y1 and Y2 could be read as L7; L1 is identical to G2, L2 to G5, G6 to L8, and so on.

Given the dialectical and diachronic similarities in pronunciation between Ancient Greek O, Y, OY, Ω (compare, for example, Attic βουλή, boulē with Doric βωλά, bōla, both meaning counsel, design, deliberation, decree), Latin U (frequently replacing the Greek O), and Celtic or English OU, OO, and the digram OW, it is conceivable that the latter is a phonetic transcription of an ancient Ω. Whereas OW does not make much graphic sense in English today, Ω does. It represents an Ω-shaped object, either convex (blown, Ω) or concave (bowl, ℧). The archaic cluster BΩ< would represent an Ω-object brought close to a B-like object, the lips. The direction of movement would be indicated by the arrow <, read as L. A bowl (BΩ<) is a drinking tool. The digraph Ω< stands for ℧, and BΩ<, for B℧. When < is on the right side of Ω, it signifies a diminution of the Ω-object, e.g., pouring from a bowl into a mouth. When < moves to the left side of Ω, it represents a magnification of the Ω-object. Between the lips (B) and the Ω-object, the grapheme < nicely describes the air blown from the lips (B) and causing the Ω-object to augment (<Ω = Ω).

In English, the cluster B<Ω reads blow, i.e., lips (B) + air stream (<) + inflated convex object (Ω). When articulated, B shows the lips while L shows the mouth cavity using the tip of the tongue. BLΩ shows the direction from the mouth cavity (BL) to the Ω-object, whereas BΩL indicates a direction from the lips (orifice) of the Ω-object to the mouth cavity (L). Whether considering their graphemes or their phonetic transcription, blow and bowl are iconic words and good examples of the iconicity of linguistic signs. The metathesis of L is not erroneous but semantically significant.

As the above hypothesis predicts, BLΩ is used for rising, expanding, or voluminous objects such as βλωρός (blōros), fig leaf, βλώσκει (blōskei), rising, and βλωθρός (blōthros), tall, well developed, also glossed as εὐαυξής (eyayxēs; Hesychius), akin to αὔξη (ayxē), dimension, and αὔξησις (ayxēsis), growth, increase, multiplication, increment, amplification, augment; compare English blather, bladder, blah, Scottish blether, Old Norse blaðra, German bladdern, French blabla, all thought to derive from Proto-Germanic *blodram, meaning to speak inarticulately, talk nonsense. In addition,  βλωμός (blōmos) is glossed as ψωμός (psōmos) and translated as morsel, bit, always referring to bread or to man’s muscles. Bread and a man’s muscles have the common characteristic of raising, and growing in volume, but a man’s muscles are never morcellated. The morsel interpretations are due to semantic drift. Bread is usually morcellated, but it is risen (blōmos) and cooked in the first place (psōmos; see Late Egyptian <bsw> and <ps> above). The original meanings of blōmos, psōmos, and Modern Greek psōmi (ψωμί; bread) were more likely raised, inflated, and baked. Instead, BΩL signifies small volumes of materials or diminished objects, e.g., βωλάζω (bōlazō), to clod, or βῶλος (bōlos), a clod of earth, lump, as of gold, nugget, hence, βώλινος (bōlinos), made of clay.

The stems bōl and blō may be confused by erroneous phonetic metathesis of consonants or vowels, as it is generally assumed, but semantic drift mechanisms should not be neglected. The French term boulanger, meaning baker, bread maker, seems phonetically closer to bōl (bowl) than to blō (blow; raised, and baked bread). How come a word about bread (Greek blōmos) starts with boul- and not with blou-? And how come the French word for bread-maker has nothing to do with the French word for full-blown and baked bread, pain? What does boulanger really mean?

The Modern French spelling, boulanger, first appears in a 1299 AD text. The Medieval Latin form with the same meaning, bolengarius, is first attested in 1100; it becomes bolengerius circa 1120, and bolengier circa 1170. It has been deduced that the first part of the word derives from the hypothetical Germanic root *bolla, as does the Dutch presumed cognate bolle, and High German bolla, all meaning round bread. According to the Germanic-origin hypothesis, the ending -anger would have derived from the hypothetical Old Frankish *-enc ending, from Germanic -ing, as in Old Picard boulenc (bun or bread maker), normalized with the French ending -ier. Thus, Old French bolengier and boulanger (baker) would have derived from Old Picard boulenc (bun-maker, bread-maker), from Low Frankish *bollā (bun) + -enc (-ing), from Frankish *-ing (-ing), from Proto-Germanic *-ingaz (-ing)[2]. This hypothesis builds upon a series of also hypothetical foundations, rests on hypothetical languages, and ignores the attested term bolengarius. The following proposition of Greek and Latin origin is much more parsimonious.

There is little doubt that the boul- of boulanger and the bole- of bolengarius are cognates of Middle French boule, meaning ball, globe, bowl, scoop, bauble, spherical object, and Old French bole, knob (edema, swell, lump, tumefaction, bump), club (stick with a thick end), mace (mass, bulk; note the antonymy of English bulk/club created by the inversion bulk/klub, *klub giving English club). In Latin, we have bulla, for any object swelling up (hence, for bread), and thus becoming round like a bubble, water-bubble, or for objects that are rounded by art, studs, amulets, metaphorically, trifle, vanity etc., and its cognates bullo, bullŭla, bulbus (or bulbŏs; Greek βολβός, bolbos), bulga, a leathern knapsack, bag, the womb, etc. Note the hollowness of bul-objects indicated by the grapheme U. In English, we have bulb, bubble, boil, and bulimia, from Latin būlīmo, from Greek βουλιμιῶ, boulimiō, to suffer from βουλιμία (boulimia), bulimia or insatiable, ravenous hunger, fill (O) and empty (U) the stomach.

Instead, Latin bu-l becomes bol- for massive spherical or semispherical objects such as bōlētus (Greek βωλίτης; bōlitēs), a mushroom, bōloe (Greek βῶλοι; bōloi; clods of earth), precious stones, bolbĭtŏn (Greek βόλβιτον; bolbiton), the dung of cattle. Therefore, the bol of bolengarius is neither hollow nor swollen but a massive spherical object. The letter U was added to boulanger either for better reproducing the /u/ phoneme of bulla (swelling object, bubble), or for a more precise graphical representation of a quasi-hemispherical object such as a loaf of bread and its mold. Yet, in French, bol primarily means clay, but retains meanings of concave objects (e.g., anus), vessels (vein), and containers (bowl, pot, jar, cup, earthenware), while its inverse, lob, is used for flat-convex objects such as lobe or the trajectory of a ball thrown to a significant height.

Here is a mechanism of semantic drift that may cause a blō-object (e.g., blōmos, bread) to be interpreted as a bōl-object (e.g., boulanger, baker), and vice versa, as announced above. The French phrase boule de glace means a scoop of ice cream. Does the term boule refer to the spherical mass of ice-crème artistically rounded by the scoop or to the scoop itself, which is a hemispherical hollow tool? Very few do care in everyday life! For some, perhaps for most of the people, boule erroneously evokes the filled, convex, sphere of ice-crème they eat, which resembles a ball (A for filling) or globe. For others, boule is the scoop resembling an empty, hollow, concave bowl (/boul/), used for measuring the amount of ice-cream to serve and, eventually, shaping it nicely. Similarly, in the context of bread-making, the French and English boule bread (Fig. 13) probably refers to the bowl used for measuring the amount of dough and for shaping the bread rather than to the shape of the bread itself.

Figure 13. Boule bread. Artwork by Zantastik. Creative Commons license.

The French bluteau, associated with Old French buretel, buletiel, blucteau, buleteau, blutoir, and thought to derive from the verb buleter or bluter + suffix -eau, is a bread-making utensil today interpreted as sieve. No precise definitions exist for those ancient terms. Their variable spelling is commonly attributed to brute dissimilation, but it may also correspond to subtle, though significant semantic differences, e.g., evolutionary variants of the tool, different uses, or downright different tools. A graphocentric theory would explain, for example, buleteau as bule + eau, giving bule-t-eau, i.e., the mixture of dough (bule) with water (French eau) – or a tool (e.g., container) for making such mixture – with the archaic (“Phoenician”) grapheme + read as T. Instead, bluteau would be blu + eau, i.e., the expansion, blow up, blow (blu, of dough) by adding (+) water (eau); or the additional rising of the dough written as blu + Ω (pronounced like French eau), and the container where it occurs. Similarly, blutoir would be the place, surface, or container (typical meanings of the French ending -toir) where the rising (blu-) occurs.

Regarding the second stem, -anger of boulanger or -engarius from bolengarius, we need not invent unattested languages and roots. In French, the E of -engarius is nasalized as /ɛ̃/, which is very close to French cardinal /a/ (French /en/ ~= Latin /an/ and Greek /ang/ from agg-). Therefore, -engarius phonetically matches the Latin angărĭus, from angărĭo, to demand something as angaria (Greek ἀγγαρία; aggaria; /angaria/; service to a lord, villeinage), to exact villeinage, from Greek ἀγγαρεύω (aggareyō; /angarevo/), to press into service, constrain, press one to serve as an ἄγγαρος (aggaros; /angaros/; mounted courier, of mules, term of abuse), probably akin to Assyrian agarru, hired laborer. In the Byzantine lexicon of Hesychius, ἀγγαῤῥία (aggarria; /angaria/) follows the Assyrian spelling with double-R and is explained as δουλεία (douleia), bondage, thrall, slavery, slave-class, hire service. The Egyptian root <Ar>, appearing in angărĭus, is written with the Gardiner hieroglyphs T12 and A24, representing a bowstring (for items that are hard, durable, strong), and a man striking with both hands (to hit or strike, power, strength, teach a lesson or instruct), respectively, and means to oppress (the poor). In Modern Greek, δουλεία (douleia; bondage, slavery) differs from δουλειά (douleia; work, labor, task, employment, business, opus, deed, doing, service) only in the tonic accent. Work may be felt as bondage when it is not justly remunerated.

Following this etymological analysis, the original and most accurate meaning of boulanger was not the noble bread maker but a dough-boule maker, a serf hired for kneading: boule + angărĭus. Graphemes, phonemes, and sememes, circulate and recombine anywhere in Egypt, Assyria, Persia, Canaan, Anatolia, Greece, Latium, and Europe. Obsessive classification of languages using Indo-European, Turkic, and Semitic family tree models may, indeed, be inappropriate, at least as far as the vocabulary is concerned. We do not know if the ’Išbaʿal inscription BOL (Ϭ𐰧Ͻ; Fig 8, characters 3, 4, and 5) was an attempt to write the Late Egyptian term <bwl> in linear European script or if the Late Egyptian had borrowed some Indo-European term created by ichnography as B℧<.’ Išbaʿal is probably a Semitic phonetic rendering of the same term for a kind of bowl, pottery material, or method, not a proper name; and baʿal, or Baʿal, is not a deity but a bowl.

In addition to ἄσβολος (asbolos; soot, literally as-bolos = fire-clay), we find sbē (/sbiː/) in Homeric ἔσβη (esbē; aorist of σβέννυμι, sbennymi), to quench, put out,  extinguish (of fire), dry out (of liquids); θίσβη (thisbē; glossed by Hesychius as σορός), a vessel for holding human remains, a cinerary urn; Modern Greek σβήνω (sbēnō), to quench, switch off, turn off, burn out, extinguish; and in English asbestos. The etymology and literal meaning of the latter term are presented as certain but may be inaccurate. Asbestos, referring to a fireproof material, is paradoxically explained as inextinguishable (Plin. Nat. 19.4), from the privative prefix a-, not, and sbestos (σβεστός), verbal adjective from sbennymi, to quench, from PIE root *(s)gwes- to quench, extinguish. To resolve the contradiction between fireproof and inextinguishable, it has been proposed that asbestos fibers were used for making inconsumable wicks (Alleman & Mossman, 1997). But inconsumable wicks are still not unquenchable.

There is a different solution to the asbestos confusion. Another word, ἀμίαντος (amiantos) has been used in Greek for the fireproof material. It is made of the privative prefix a- and μιαντός (miantos; stained, defiled), from the verb μιαίνω (miainō; to stain, sully, taint, defile; see section Cadmus and Cilix) and means undefiled, pure, free from stain, not to be defiled. As an adjective of λίθος (lithos; stone), amiantos lithos specifies asbestos. By analogy, asbestos is always split as a-sbestos, thinking of a privative a- and the verb sbennymi (to quench). Perhaps, this splitting is wrong.

With an archaic Asb meaning glowing flame and ἕννυμι (‘ennymi) meaning to put clothes on, cover, wrap, shroud, wear, Asb-‘ennymi could have adapted to Greek grammar as sb-ennymi (sbennymi; to quench), originally meaning to cover fire with cloths. Among various Homeric aorist forms of ‘ennymi, we find ἕστο (‘esto, ϝesto, vesto or hesto; Il.23.67). It compares to Latin vestis (the coverings for the body, clothes, garments clothing, vesture, carpet, curtain, tapestry, any sort of covering or coating), Sanskrit váste (clothes himself), and Greek βεστίον (vestion) and βέστον (= ἱμάτιον; veston), both meaning clothing, a piece of dress, an outer garment, generally, clothes. The stem esto or besto (pronounced /vesto/) is found in all, ἄσβεστος (asbestos) glossed as not quenched, also unslaked lime (calcium oxide), plaster, ἀσβέστινον (asbestinon), non-combustible material, Modern Greek ασβέστης (asbestēs), unslaked lime, quicklime or burnt lime, ἀσβέστιον (asbestion), calcium, and English asbestos. Unslaked lime, quicklime, or burnt lime (Greek asbestos or asbestēs) is produced by the thermal decomposition of materials, such as limestone or seashells, that contain calcium carbonate, by heating the material in a lime kiln to above 825 °C (1,517 °F)[3]. The burnt sememe resides in the asb (or Egyptian <Asb>) stem of asbestos and its cognates. Thus, asb-est-os reads as the fire-cloth, fire-cover, or fire-coat thing.

Evidently, these stems form a basis with fire-protection semantics which may be ichnographically elaborated to modulate the meaning of derivative terms. Fire protection may mean protection of the fire or protection from the fire. A covered fire is protected from rain and strong winds which may extinguish it. A well covered and protected fire is relatively unquenchable (asbestos), but the Greek adjective asb-est-os literally means protected fire, not unquenchable (a-sbestos). The English asbestos reuses the stems to name a material that protects from fire, a fire-proof coat or cloth. The unslaked lime is a burned (asb) coating (est) thing (os), asb-est-os. Using the same term for a well-protected fire and a fired material will inevitably cause confusion. The Ancient Greek term asbestos for burnt lime evolved to Modern Greek asbestēs, where the confinement of a protected fire, evoked by a circle (O), was replaced with the sememe of a large surface evoked by H (ē). Firing renders some materials, such as clay (bol-os), fireproof. Fired earthenware resists cooking conditions, whereas raw clay vessels may break. A fired, fireproof clay vessel could well be called asb-bol, contracted as asbol (fired bowl). To extinguish a fire, we normally stop feeding it. The letter A is interpreted as filling throughout in this series of essays.

Similarly, in Asb, i.e., glowing, radiant (fire; Egyptian Asbyw), A provides the sememe of filling in the sense of feeding the fire. By removing the A from Asb, we essentially stop feeding the fire. We thus finish with a sb-est, as in sbestos (σβεστός), meaning quenched, extinguished, i.e., not fed.

The crosstalk between Egyptian and Greek also provides an elegant secular interpretation of the rest of the ’Išbaʿal inscription …IBSIBDO. After the invention of Greek Ψ (Psi; /ps/), the cluster IBS would be transliterated in Greek as IΨ. The Homeric ἴψ (ips) has been translated as a borer, a worm that eats wood and opens holes. Most likely, the animal was named borer after its characteristic behavior of entering a solid material and leaving a hole behind or blocking a hole (Table 1). A tool for opening holes could have been named after ips, but ips may also create names for objects that fit into holes, like a stopper or a cork (ἴψος; ipsos) The inverse stem, <bsi>, means flow forth of water in Middle Egyptian (Dickson, 2006). It retains similar meanings of flow, coming forth, bringing forth, or emerging in Late Egyptian, where it also gains meanings like introduction, induction, admittance into, initiation, or installation (Lesko, 2002). Inversion of the Greek ἴψ (ips) gives ψῖ (psi), the name of the Greek letter Ψ. Like most Greek letter names, Psi is considered a meaningless proper name. We may, however, deduce its meaning by semantic pairing. Paired with the simplest suffix (-ω) to form a verb, ψίω (psiō), or with -as, to form a noun, ψιάς (psias), psi reveals its sememes of controlled flow from a hole of a container for feeding or drinking, like a breast’s nipple. With this notion in mind, we can appreciate Ψ as a pictogram of a container perforated with a stick. The I, preceding or following Ψ, determines whether the perforation of this Ψ-container is blocked by the object (ιψ; ips; borer, cork) or flow is allowed (ψῖ; psi; feed, drip, drink). Evidently, there is some phonetic and semantic relation between the Greek psi and the Egyptian <bsi>, but the meanings of active perforation and passive flow are confounded in the same Egyptian word and the inversion does not work the same way, if at all.

It is possible that some of these roots were somewhat misinterpreted by modern translators of either Egyptian or Greek. Alternatively, Egyptian writing was not as rigorous as Greek. For example, the metathesis of I from <isb> to <sbi> has no semantic consequences in Egyptian since both stems mean vanish. The Egyptian <bsi> is comparable to Greek stem psi, both related to flow, but the inversion ibs creates an antonym of flow only in Greek; <ibs> is unrelated to flow in Egyptian. Instead, the Late Egyptian flow words, <bsi> and <sib>, derive from Middle Egyptian <sbi>, drink, by reshuffling metathesis of I or B. In Late Egyptian, <sib> has many seemingly unrelated meanings, among which we find the morphologically similar English terms flower, as well as jackal, judge, joy, and javelin – all starting with the extremely rare J – suggesting an iconic relation of J with <sib>. A javelin is also semantically related to English flow since, of solids, to flow means to undergo a permanent change of shape under stress, without melting. As a javelin, <sib> would describe the method of manufacturing such objects.

Greek uses sib, supplemented with -ύνη (-ynē), to form σιβύνη (sibynē) for objects similar to a javelin like a spear or a pike. The stem ynē is a concatenation of yn, from ὐνω (ynō; to make thin, pare away, fine down, grind small; or become thin, watery, of a fluid), with H (ē) for length. The antonymy produced by inverting SIB into BIS can be sensed in Fig. 14, comparing sibynē (spear, pike) to bis-derivatives like βίσβη (bisbē), a pruning hook, or βίσων (bisōn), bison. Mainstream phonocentric theory suggests that bison is ultimately of Baltic or Slavic origin (the existance of the Ancient Greek cognate remains unnoticed), and means the stinking animal in reference to its scent while rutting.

Figure 14. Objects named after the inverted Greek stems sib and bis. The sib-objects (from sibynē: A=spear, B=pike) are long, linear, sharp 'killing tools' (weapons). The bis-objects are short, curved, sharp 'killing tools' like the pruning hooks (bisbē; C and D) or the horns of a bison (E).

When supplemented with the particle δή (dē; in truth, indeed, surely, really, quite, verily, like, so, this and no other, above all, plainly, very, only) to provide further exactness, sib gives σίβδη (sibdē), flow, flux. Metathesis of Sigma from the beginning of sibdē to the end of ibdēs (ἴβδης) creates an antonym of flow (sibdē), the flow-stopper (ibdēs; cock, plug). An object represented by Sigma seems to change position from front to back to allow flow or stop it. Once more, semantics correlate with morphology in Greek but not so in Egyptian. It may be proposed that Greek borrowed Egyptian roots and organized them into an improved, systematic writing system whereby graphemes and word morphology suggest a meaning. Alternatively, Late Egyptian randomly borrowed Greek constructs ignoring the rules under which these were created.

Finally, the cluster BD (lips + passage) is combined with A (filling) to give BDA as in βδάλλω (bdallō), to milk, suck, and with E (opening) for βδέω (bdeō), to break wind. Thus, BDA is for in-flow and BDE for out-flow through an orifice.

This analysis suggests that the ’Išbaʿal inscription uses Greek roots, some borrowed in Late Egyptian with similar semantics. In Old European/Archaic Greek script, reading ASBOLIBSIBDO, the inscription may mean something like ‘fired bowl [with] bottom-flow control [tap, emptying hole and plug]’.

Table 1. Semantic analysis of IBSIBDO.

IBSIBDO	EGYPTIAN	GREEK
IBS (IΨ)	<ibs> headdress (Middle Egyptian) <ibi> be thirsty (Middle Egyptian)	ιψ (ips) ἴψος (ipsos)	wood-worm borer cork
BSI (ΨI)	<bsi> flow, flow forth (of water), influx, introduce, induct, emerge, admit into, initiate, install	ψῖ (psi) ψίω (psiō)   ψιάς (psias)	letter Ψ feed on pap,  give to drink drop
ISB	<isb> = <sbi> vanish
SBI	<sbi> drink (Middle Egyptian), go, travel, attain, watch over, send, conduct, spend, pass, attain, approach, be faint, perish, vanish <ssbi> despatch (deprive?)
SIB	<sib> flow, jackal, dignitary, judge, speckled snakes, destroy, flower, joy, javelin, sloth	σιβύνη (sibynē)	spear, pike
SIBD	<Sibd> stock	σίβδη (sibdē)	flow, flux
SIBDO		στιβδός (sibdos)	pricked, tattooed, dappled
IBD	<Ibd> month	ἴβδης (ibdēs)	cock or plug in a ship’s bottom
IBDO	<Ibdw> fish
BD	natron, glaze	βδάλλω (bdallō) βδέω (bdeō)	to milk, suck   to break wind

Claims

The frequencies of letters, or letter clusters, may establish graphocentric linguistic phylogenies or identify a language behind an undeciphered script.

The Ishbaal inscription may be read as Greek, meaning ashtray (ash-bin, dustbin), soot-ink decorated, or fired bowl with bottom-flow control [tap].

Cognates

Ishbaal (Greek root AShBOL-): English ash.

BOL: English bowl.

 

References

Ager, S., & Paliga, S. (2021). Vinča symbols. Omniglot.

Alleman, J. E., & Mossman, B. T. (1997). Asbestos Revisited. Scientific American, 277(1), 70–75. 

Augustyn, A. (2020). Philistine. In The Editors of Encyclopaedia (Ed.), Encyclopedia Britannica. britannica.com. 

Beekes, R. S. P. (2014). Pre-Greek: phonology, morphology, lexicon. Brill.

Beekes, R. S. P., & Beek, L. van. (2010). Etymological dictionary of Greek. Brill.

Benz, F. (1972). Personal names in the Phoenician and Punic inscriptions : a catalog, grammatical study, and glossary of elements. Biblical Institute Press.

Coleman, J. E. (2000). An Archaeological Scenario for the “Coming of the Greeks” ca. 3200 B.C. The Journal of Indo-European Studies, 28(1–2), 101–153. 

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

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

Drinka, B. (2013). Phylogenetic and areal models of Indo-European relatedness: The role of contact in reconstruction. Journal of Language Contact, 6(2), 379–410. 

Dunand, M. (1945). Byblia Grammata. Documents et recherches sur le développement de l’écriture en Phénicie, Beirut (Vol. 2). Libanese Ministry of National Education and Fine Arts.

Dunand, M. (1978). Nouvelles inscriptions pseudo-hiéroglyphiques découvertes à Byblos. Bulletin Du Musée de Beyrouth, 30, 51–59.

Dussaud, R. (1924). Les inscriptions phéniciennes du tombeau d’Ahiram, roi de Byblos. Syria, 5(2), 135–157. 

Everson, M. (2020). Final proposal to encode the Cypro-Minoan script in the SMP of the UCS (No. N5135). International Organization for Standardization.

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. 

François, A. (2015). Trees, waves and linkages: Models of language diversification. In C. Bowern & B. Evans (Eds.), The Routledge Handbook of Historical Linguistics (pp. 161–189). 

Garfinkel, Y., Ganor, S., & Hasel, M. G. (2018). Introduction. In Martin G Klingbeil (Ed.), Khirbet Qeiyafa Excavation Report 2007-2013 Vol. 4: Art, Cult, and Epigraphy (Vol. 4, pp. 3–13). Printiv.

Garfinkel, Y., Golub, M. R., Misgav, H., & Ganor, S. (2015). The ʾIšbaʿal Inscription from Khirbet Qeiyafa. Bulletin of the American Schools of Oriental Research, 373, 217–233.

Gimbutas, M. (1986). Remarks on the ethnogenesis of the Indo-Europeans in Europe. In W. Bernhard & A. Kandler-Palsson (Eds.), Ethnogenese europäischer Völker (pp. 5–19). Gustav Fische Verlag.

Gimbutas, M. (1997). The Kurgan culture and the Indo-Europeanization of Europe: Selected articles from 1952 to 1993 (Vol. 18). Institute for the Study of Man.

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

Golub, M. (2014). The distribution of personal names in the land of Israel and Transjordan during the Iron II period. Journal of the American Oriental Society, 134(4), 621–642.

Lesko, L. H. (2002). A dictionary of Late Egyptian. B.C. Scribe Publications. 

Levin, Y. (2018). The Identification of Khirbet Qeiyafa: A New Suggestion. Bulletin of the American Schools of Oriental Research, 367, 73–86.

Martirosyan, H. (n.d.). Correlations in the origin of Armenian pictograms and the ancient world writing systems 1. The Egyptian hieroglyphs. Academia.Edu. Retrieved 25 February 2022.

Naveh, J. (1978). Some Considerations on the ostracon from ʿIzbet Ṣarṭah. Israel Exploration Journal, 28, 31–35.

Petrolito, T., Petrolito, R., Winterstein, G., & Perono-Cacciafoco, F. (2015). Minoan linguistic resources: The Linear A Digital Corpus. Proceedings of the 9th SIGHUM Workshop on Language Technology for Cultural Heritage, Social Sciences, and Humanities, 95–104.

Renfrew, C. (1973). Problems in the general correlation of archaeological and linguistic strata in prehistoric Greece: the model of autochthonous origin. In R. A. Crossland & A. Birchall (Eds.), Bronze Age migrations in the Aegean; Archaeological and Linguistic Problems in Greek Prehistory: Proceedings of the First International Colloquium on Aegean Prehistory, Sheffield (pp. 263–276). Duckworth.

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

Sass, B. (2019). The Pseudo-Hieroglyphic Inscriptions from Byblos, Their Elusive Dating, and Their Affinities with the Early Phoenician Inscriptions. In Par la bêche et le stylet! Cultures et sociétés syro-mésopotamiennes (pp. 157–180). Archaeopress Publishing Ltd. 

Sass, B., Garfinkel, Y., Hasel, M. G., & Klingbeil, M. G. (2015). The Lachish Jar Sherd: An Early Alphabetic Inscription Discovered in 2014. Bulletin of the American Schools of Oriental Research, 374, 233–245.

Sass, Benjamin. (1988). The genesis of the alphabet and its development in the second millennium BC. Harrassowitz.

Tsuen-Hsuin, T. (1985). Paper and Printing. In J. Needham (Ed.), Science and Civilisation in China: Chemistry and Chemical Technology (Vol. 5). Cambridge University Press.

Vahanyan, G., & Vahanyan, V. (2009). The Intercultural relations between Old Europe and Old Armenia. Papers, XXIII Valcamonica Symposium, Prehistoric and Tribal Art: Making History of Prehistory, the Role of Rock Art, 357–362.

Vita, J.-P., & Zamora, J.-Á. (2018). The Byblos script. In S. Ferrara & M. Valério (Eds.), Paths into Script Formation in the Ancient Mediterranean (pp. 75–102). Edizioni Quasar.

Winn, S. M. M. (1981). Pre-Writing In Southeastern Europe: The Sign System Of The Vinča Culture Ca. 4000 BC. Western Publishers.

 

---

[1] Amarna letters in English Wikipedia. Accessed 23 August 2021.

[2] Byblos Necropolis graffito in English Wikipedia; accessed 30 June 2021.

[3] Instances of the AB47 sign in the SigLA database; accessed 30 June 2021.

[1] In the American Heritage Dictionary, the only Semitic word ending in -yw, is <ḥyw>, to live. Instead, there are many Ancient Egyptian words ending in -yw. With very few exceptions, they describe or refer to people, i.e., objects that rise and fall, live and die. From Dickson (2006) we gather: <aDyw>, winnowers; <aHwtyw>, cultivators; <antyw>, myrrh; <artyw>, they who ascend; <Asbyw>, flames; <Axtyw>, horizon - dwellers  (a remote people; <bAgyw>, the languid ones (the dead); <bnrytyw>, confectioners; <bSttyw>, rebels; <bwytyw>, those who are abominated; <DAytyw>, opponents; <DAyw>, opponent; <Drtyw>, ancestors; <dwAtyw>, dwellers in the netherworld; <HAyw>, carrion - birds; <Hmsyw>, guests; <Hmwtyw>, craftsmen; <Hnsktyw>, wearers of the side-lock (of hair); <Htptyw>, the peaceful ones (the blessed dead); <Htpyw>, non-combatants; <iAbtyw>, Easterners; <iAtyw>, mutilation; <imAxyw>, revered ones (of the aged living); <imntyw>, Westerners; <iryw>, crew (of boat); <iwntyw>, tribesmen; <iwtyw>, corruption; <kftyw>, (locality) Crete ?; <knmtyw>, they who dwell in darkness (name of a conquered people); <mAatyw>, just men, the righteous (the blessed dead); <mabAyw>, the Thirty (a judicial body); <mHtyw>, northerners; <msTyw>, offspring; <myw>, semen, seed of man; <nDtyw>, maidservants ?; <niwtyw>, citizens, townsmen; <nnyw>, inert ones (the Dead); <nsyw>, Kings; <pAwtyw>, men of ancient families; <pDtyw>, foreigners; <pwntyw>, the people of Punt; <r pDtyw>, foreigners; <rmnwtyw>, (pl.) companions; <rsyw>, Southerners; <sDAwtyw>, treasurers; <sdtyw>, weaklings; <smntyw>, emissaries; <smytyw>, owners of a desert tomb; <spAtyw>, nome-men; <sTtyw>, Asiatic; <styw>, Asiatic, Nubians; <sxryw>, those who govern; <wHAtyw>, oasis - dwellers; <xAstyw>, foreigners, desert-dwellers; <xbstyw>, bearded ones; <Xnwtyw>, skin - clad people; <xrwyw>, war.

[2] See also boulanger in Wiktionary. Accessed 16 June 2022.

[3] Calcium oxide in English Wikipedia; accessed 19 June 2022.