THE ORIGINS OF THE PROTOALPHABET
Brian E. Colless
Note that this essay has been superseded
by an expanded published article:
The Origin of the Alphabet:
An Examination of the Goldwasser Hypothesis
Antiguo Oriente, volumen 12, 2014, pp. 71–104.
Note that this essay has been superseded
by an expanded published article:
The Origin of the Alphabet:
An Examination of the Goldwasser Hypothesis
Antiguo Oriente, volumen 12, 2014, pp. 71–104.
Since 2006 the discussion of the origin of the Semitic alphabet has been given an impetus through a hypothesis propagated by Orly Goldwasser: the alphabet was invented in the 19th century B.C.E. by illiterate Semitic workers in the Egyptian turquoise mines of Sinai; they saw the picturesque Egyptian inscriptions on the site and borrowed a number of the hieroglyphs to write their own language, using a supposedly new method which is now known by the technical term acrophony. Twenty-one propositions from Goldwasser’s publications are examined critically here.
In 1916 Alan Gardiner published his theory on the origin of the alphabet, plausibly proposing that the original letters of the alphabet, as represented in the Semitic inscriptions from the turquoise mines of Sinai, were borrowed from the Egyptian store of hieroglyphs.1
In 2006 Orly Goldwasser went further, arguing that the Semitic workers at the Sinai mines had actually invented the alphabet to write their own language, employing the Egyptian pictorial signs they saw in the stela inscriptions at the site, but with no understanding of the Egyptian writing, and in the process they created the system of acrophony. 2
“I believe that the inventors of the alphabet did not know how to read Egyptian. When they looked at the Egyptian sign [N35] (N in Egyptian) they recognized the picture of water. In Canaanite (their language) the word ‘water’ might have been mem or maim. From this word they took the first sound alone – M; which became the letter mem in the Canaanite scripts, and finally the English letter M.” Orly Goldwasser. 3
We will here examine twenty-one points made by Goldwasser in the presentation of her thesis,4 and confront them with findings from my own research on the same subject.5
1. The alphabet was invented in Sinai, before year 13 of Amenemhet III
This means that the invention (or innovation) took place in the nineteenth century around 1840 B.C.E. It is true that the protoalphabet is represented in inscriptions in Sinai in the Middle Kingdom period, but there is no explicit evidence that it was invented there rather than elsewhere. A scribal school in some city of Canaan or Egypt is a more likely setting, but the place and time are still a mystery. Nevertheless, it is quite possible that the oldest-known protoalphabetic inscription has been found in Sinai (perhaps 349=22).6
2. All hieroglyphic prototypes for the letters of the alphabet clearly exist in the hieroglyphs of the Sinai inscriptions of this period
This assertion may be true (excepting the signs that do not have an Egyptian model, namely W, T, T, in my view) but it is not “clearly” substantiated by Goldwasser, as her tables of protoalphabetic signs with presumed Egyptian counterparts (“Graphemes of the Protosinaitic Script”) have only 21 items (though she knows it had several more).7 Her identifications of half of the signs on her table are faulty, in my view. Het and Tet are not registered, but they occur in the Semitic texts, and their prototypes are present in the Egyptian inscriptions at the site. She singles out the letter He as a “special link”: it depicts a man standing with his arms raised (each forming a right angle, at either side of the head) and it will ultimately be the Greek and Roman letter E; in her opinion the sign (a version of hieroglyph A28) probably represents a local title related to the expeditions, since it is rare in Egypt; but she has chosen the wrong hieroglyphic examples for the origin of He (see my discussion in section 12 below).
3. About 30 Protosinaitic inscriptions were found in the area of Serabit el-Khadem
It should be added that one (Sinai 348=9) was found in the Wadi Magharah area but it is now lost, though two copies have survived; and my tally is 44, not merely 30. The considerable number of items is used as a weak argument for the origin of the protoalphabet in Sinai: “The only reasonable explanation for such a ‘boom’ in this kind of writing in Sinai is that Sinai was the site of its invention.” 8 However, another way of looking at it is that stone was readily available at this site, for writing Egyptian or West Asian inscriptions, and they have survived, albeit damaged by weathering, though not destroyed (as texts on perishable material would have been).
4. All but one of the texts show very early paleographical stages of the script, and were probably produced during a rather short span of time
The one exception she adduces is “Sinai 375c” (381=41 in my numbering scheme), attributed to the New Kingdom; but others must also be NK. Gordon Hamilton has two main categories in his chronological scheme for the Sinai texts: earliest (ca 1850-1700), and typologically developed (ca 1700-1500) 9; by this criterion Goldwasser’s “short span” is impossible, needing to encompass the centuries from the 19th to the 16th; but certainly there are numerous examples from the 12th Dynasty (Middle Kingdom, 19th C) and also the 18th Dynasty (New Kingdom, 16th C); and the pictorial aspect is constant throughout.
5. In Egypt only three such inscriptions are known
To arrive at this total, Goldwasser takes the Wadi el-Hol inscription as two items (but it is a single continuous text, in my reading of it) 10 and she adds the ostracon from the Valley of the Queens (see section 12 below). There are more than three protoalphabetic inscriptions, plus several syllabic texts; the three most important of them have copies of the letters of the protoalphabet on them (they are not texts), and they are from the New Kingdom period.11
6. Only a handful of early alphabetic inscriptions are known from Canaan, scattered along the Late Bronze Age, from the 18th to the 13th century B.C.E.
Actually, besides fragmentary and minimal texts, there are more than a dozen that make a statement, all on non-perishable materials.12 None are extant on papyrus or parchment, though such must have existed; absence of evidence is not evidence of absence; in the Iron Age the situation is similar, since the writings of the Phoenicians are lost, but we know they kept records (see also sections 19 and 20 below).
Goldwasser’s supposition is that the protoalphabet was not widely used, and was the lowly person’s writing system; but the examples from Canaan are from temples (on paraphernalia such as bowls, an incense stand, and a ewer), from factories (on pottery), and from a tomb that did not belong to a pauper (on a dagger).
7. The inventors were illiterate, that is, they could not read or write Egyptian
Of course, when they had invented the new script for writing their own West Semitic language they would have become literate. But Goldwasser is assuming that they had no knowledge or understanding of writing, including the Egyptian system. As we progress through her propositions we will see reasons for doubting this claim; a notable example is the use of the nefer hieroglyph (F35) for Tet in the protoalphabet, an icon that is rather opaque and not self-explanatory (see section 15 below).
8. This illiteracy motivated them to formulate new relations of sound and icon, and to come up with a new solution, namely acrophony
In fact the acrophonic principle was already established for them in their West Semitic syllabary, dating back to the Old Kingdom in the Early Bronze Age.13 This significant fact alone should invalidate Goldwasser’s basic hypothesis. The acrophony (“summit sound”) principle was arguably a modification of the rebus principle, and it was the mechanism for constructing the syllabary and subsequently the consonantary.14 In the syllabary, a picture of a door evoked the word daltu (door) and acrophonically this yielded the syllable DA; in the consonantary (the protoalphabet) it was D. In one case there was a monosyllabic word: the picture of a mouth said pu and then P.
9. The Sinai protoalphabetic inscriptions consistently show the wrong direction of writing according to Egyptian rules
This is the first of a number of alleged indications of the illiteracy of the inventor or inventors of the protoalphabet.
Egyptian scribes arranged their texts either in blocks of horizontal lines or in sets of vertical columns, and in both cases the preferred direction was from right to left. With regard to horizontal lines of writing, Egyptian scribal practice was to run the characters from right to left (sinistrograde, as in Hebrew and Arabic), but sometimes from left to right (dextrograde, as in English writing). Egyptian signs with fronts and backs (such as birds and human heads) must look backwards to the beginning of a horizontal line of writing. Most of these Semitic inscriptions run down in columns; only one of them has a block of inscribed lines: Sinai 349=22 (originally an immaculate member of the corpus, but now mutilated by weathering) has the bovine and human heads facing towards the end of the line (that is, they are looking where they are going); the direction of writing on 349 is sinistrograde (running from right to left), and this is in agreement with the Egyptian tradition, but the animals (ox, snake, human) are facing the wrong way. Inscription 357=32 has a vertical column leading into a dextrograde line (left to right); the latter has the heads facing rightwards; this is opposite to Egyptian practice.
When an Egyptian text is arranged in vertical columns (with the text running from top to bottom in each line, and the columns moving from right to left) fronts and faces are regularly turned to the right; but in the previously mentioned vertical section of protoalphabetic 357=32 the oxen, the snakes, and the fishes look leftwards, as do the animals on the two columns of 351=23 (with a drawing of the god Ptah facing leftwards, the opposite of the depiction of Ptah on Sinai Egyptian Stela 92, a document which is central to Goldwasser’s thesis). These three texts (349, 351, 357) are from Mine L; they are carefully inscribed and may be some of the oldest on the site. They are deviant according to Egyptian rules, but here is the important point: they are actually faithful to their own West Semitic tradition of writing, which goes back to the time of the Old Kingdom; the syllabic texts from Gubla (Byblos on the coast of Lebanon) have bees and birds looking left along sinstrograde lines (Texts A, C, D) and looking left in columns (Text G).15 Note that there is at least one syllabic inscription from the turquoise region of Sinai (526); it moves from left to right, and the signs face that way, contrary to the Egyptian style.16
What we have in these significant inscriptions is the work of scribes who were apparently well acquainted with their own two writing systems (syllabary and consonantary), and it is likely that they were always, on successive turquoise mining ventures, bringing the protoalphabet with them from elsewhere, rather than inventing it on the spot on one of these expeditions.
However, absolute consistency in orientation (note the word “consistently” in Goldwasser’s statement) is hard to find in the rest of the collection: on the block statuette from the temple (346=4), the one fish and the two human heads look one way while the three snakes face in the opposite direction; the fish and the ox on 363=16, 352=28, and 358=35 are looking in different directions. Moreover, the heads on 356=29 (from Mine L) are actually in conformity with the Egyptian convention (whether by accident or design), as is the writing in the two horizontal lines on the bilingual sphinx (345=3). Some of these cited inscriptions would be from the New Kingdom period, presumably, and would not be relevant to the question of the beginnings of the protoalphabet.
In summary, the writers of the proto-alphabetic inscriptions (as also the syllabic texts) show a fairly consistent tendency to observe their own long-standing custom of writing from right to left, which was possibly something they imitated from Egyptian practice; but they usually had letters facing forward, in the direction of the line, which is (whether accidentally or intentionally) the opposite of Egyptian practice.
10. Letters in one and the same inscription may show different stances
As examples Goldwasser cites 357 (inside Mine L) and 358 (inside Mine M, which is joined to Mine L), both of which have already been mentioned here; in each case, the two instances of the letter L are in a lying position on the one hand and standing obliquely on the other; this is an interesting observation; a possible explanation is that the two instances are considered to be in different sorts of lines (horizontal versus vertical). In the Wadi el-Hol inscription the only L (at the bottom of the column) is inverted, and this makes a total of five variant stances for that letter. No such variety would be permitted in the work of a trained Egyptian scribe, but the Semitic writers are simply showing their individuality, not their ignorance.
In later times (Iron Age I) the different stances of letters were significant: a new three-vowel syllabary was created in Israel, apparently.
11. In most cases, the writers do not follow any order in writing
Again, Goldwasser’s generalizing is too sweeping. Her prize trophy is the jumble of letters on one side of the block statuette (346=4): the line of writing meanders. “No Egyptian scribe would ever produce such an inscription”. But it alerts us to the need to take meandering and clustering into account when reading some of the inscriptions: 365=8, 363=16, 361=13. Two examples of the line taking a sharp turn are 357=32 and 358=35); each begins as a vertical column and then runs sideways. This is untidy, but sometimes the surface of the wall might be to blame for such irregularities. Also, mining tools were used for the task; this is stated in 376=1; it is on a rock face in the oasis where water for the expeditions was obtained; its four columns are in boustrophedon style (“as the ox ploughs”, 1 down, 2 up, 3 down, 4 up) starting on the left; a large fish is standing on its tail to fit into column 3 (certainly a “different” stance); the bovine and human heads are facing right so we know which direction the text is taking (left to right). These apparent anomalies have their own artistry and attractiveness (like Arabic calligraphy), and generally the writers have aimed at setting their texts down in orderly lines or columns.
12. Two different hieroglyphs may be used as prototypes for a single letter
Goldwasser rightly adduces N, which is found as a viper (I9, f) and as a cobra (I10, d); but she does not show an example of the viper; the ideal cases for this purpose are the two similar inscriptions 360=14 (Mine K) which has a viper, and 361=13 (Mine N) which has a cobra in the corresponding position in the sequence of signs.
R is a human head, and she suggests that it is based not only on the profile hieroglyph (D1 tp) but also the frontal form (D2 Hr), which is perhaps evidenced on 364=37, 365=8, and 367=17.
H (a man with upraised forearms, the forerunner of the letter E in the alphabet) could be cited in this connection: Goldwasser has assumed that this is a high officer shouting Hey or Hoy, but the more likely link is a person exulting or jubilating (hillul celebration, already used in the syllabary for HI). A28 (man with both forearms raised) is more refined (and clothed) than the simple stick figures in the Semitic texts, but as the determinative marker for high and joy it provided the semantic basis for HI and H in the West Semitic scripts. The character for H is also found in an inverted stance (standing on hands, as in hieroglyph A29) at the end of 358=35. It occurs on the Valley of the Queens ostracon in Egypt;17 it could be taken as an inverted K, but as H it produces the word ’mht (maidservants), and this goes with the reading of the word below it as ’sht (women, though ’tt is the expected form in the Bronze Age) with the letter Shin as the NK form of the sun with a single uraeus serpent and a tail (N6). Furthermore, on the Wadi el-Hol inscription in Egypt, one of the three instances of H (fifth in the vertical line) has only one forearm raised, and Goldwasser tries A1 (seated man) and A17 (seated child with hand to mouth); but the man is obviously dancing for joy, as with hieroglyph A32, likewise denoting jubilation (but if he is kneeling, then A8 could be invoked, being yet another jubilation sign).
Rather than demonstrating ignorance of the Egyptian system, this evidence indicates knowledge of its contents on the part of the Semitic users of the protoalphabet.
13. In some cases new iconic readings are given to Egyptian signs
This statement is baseless, in my view. Goldwasser makes great play with possible origins for the letter Waw (oar, mace, toggle-pin) but disregards the fact that waw means “hook” or “nail” (WA in the syllabary, W in the consonantary); in the protoalphabet it was a circle on a stem, and then in the Phoenician alphabet the circle was opened up at the top, and this was an inverted version of the form it had in the syllabary. This object had no counterpart among the hieroglyphs, and so it leads into the next point.
14. Some letters have their origin outside of the Egyptian hieroglyphic system
Referring to my table of signs to see the forms (and notice the BS column for the syllabic forerunners attested at Byblos, and elsewhere):
W (waw) would fit here, as just noted.
T (taw, + or x, meaning a mark or signature) has no Egyptian counterpart.
T (tad breast) in frontal pose has no Egyptian prototype.
Goldwasser puts two signs into this category: hand and bow. The bow (composite bow, *tann, the supposed source of T and/or Sh) is a phantom, seen only in speculation. 18
K as a hand finds little to compare with hieroglyph D46 (but fingers can be shown in this character, to compare with the upright hand on 349=22; but it has two different precursors in the syllabary (see section 15).
15. Lack of standardization
B With regard to the letter Bet, Goldwasser declares that there is a plethora of house forms in the Sinai Semitic inscriptions, but she goes too far when she includes houses with multiple rooms, such as the one with two rooms and a rounded courtyard in 380=11 at Mine G; this is actually the sign for Het.
H. Other examples of Het are: the two-part dwelling in text 360=14 at Mine K, and the three-section mansion in 361=13 at Mine N. These are based on Canaanian mansions, the relevant word being Haçir, and comparing this to hieroglyph O6 (mansion) is legitimate, given the similarity to the Semitic forms in examples in Sinai Egyptian inscription 28.19
H The original hieroglyph (V28) is one of the single-consonant signs, representing Egyptian h. but Semitic h. It is a hank of thread, or a wick, shaped like a double helix. Sinai inscription 53 (dating from year 44 of Amenemhet III) 20 is a splendid piece of Egyptian calligraphy (and it has a host of hieroglyphs that are prototypes for the letters of the alphabet, though eleven of them are missing by my calculation); but it shows inconsistency, having some cases of this sign with the standard three loops (lines 1,4,7) and others with only two loops (lines 4,8). In the Sinai protoalphabetic texts the form with three loops is found once (376=1) and the remaining few instances have two loops.
Likewise on Sinai 53, hieroglyph F35 (nfr) shows variation: it has one stroke at the top of its stem in lines 1 and 15, and two strokes in line 6. Its equivalent (as Tet) in Sinai 351=23 has only one stroke, though in the syllabary it could have two strokes (see the BS column on my table of signs).
T. Tet only appears once in the Sinai corpus, in 351=23, but it was clearly a borrowing of the Egyptian nefer hieroglyph, which stands for goodness and beauty, and Semitic t.ab (good, beautiful) provides the acrophone (in the syllabary and the consonantary: T.A and T.). As this hieroglyph does not readily yield up its meaning (the heart and the windpipe perhaps expressing emotional reaction to goodness) knowledge of the Egyptian symbol would be required.
16. Cases of two icons competing for the representation of a particular sound
Again the accusation is lack of standardization.
D Goldwasser sets up a false opposition for the letter D: dalt (door, which is still obvious in the Roman form of D) versus dag (fish), though she knows that the fish sign could be (and surely is, I would say) Samek.21 So there are two origins for S (samk): fish and spinal column (see the table of signs). The fish occurs only in the south, apparently; the northern form is the column, as also in the syllabary.
Two more instances can be offered; both come out of the syllabary.
K kapp (palm of the hand, KA) and kipp (palm branch, KI).
M horizontal waves (water, MU) and vertical wavy line (rain, MI).
These three examples prove that the formation of the protoalphabet was basically a matter of choosing phonograms from the Canaanian syllabary, and using them as consonantograms.22
17. The acrophonic script shows no signs of contamination from the complex Egyptian ideographic system
Goldwasser asserts that “its complicated semiotic mechanism escaped them”. On the contrary, its workings had been understood by their scribes since the time of the Old Kingdom, when the West Semitic logosyllabary was constructed. In the new acrophonic script, the protoalphabet, the logoconsonantary (possibly conceived because the Egyptian system was already grasped and known) each sign could also function as a logogram (for example, house icon as bayt), and as a rebogram (the consonants of the word that goes with the picture could be employed to express homophones, or act as components in another word). Examples have been collected and presented elsewhere.23 Such extended usage of the letters seems to have been still operating in the early Iron Age, in the text of the Izbet Sartah ostracon.24 It is detectable in the Wadi el-Hol text.25 Strangely, Goldwasser allows a “classifier” sign into this inscription (letter 5 on the vertical line);26 in my interpretation it is a dancing man and a logogram (hillul “celebration”); she wants it to be a human male classifier preceding a male personal name, but the name following it is ‘Anat, the goddess who is pictured beside the name.
18. No clear case of borrowing from the monoconsonantal repertoire
This is true, but eight of them do turn up in the inscriptions, with different sound-values, not having the original Egyptian sounds: B (h) H (h.) Y (‘) K (d) M (n) N (f) N (d) P (r).
19. No hieratic signs mixed with hieroglyphs
If stylized “hieratic” signs (in which the original image is almost obliterated) were borrowed for the protoalphabet, this would require literacy as a precondition, and she insists that the inventors “interacted only with the pictorial meanings of the signs”.27 However, others disagree with her, she acknowledges.
Obviously, Goldwasser would have no time for the discredited idea that the letters of the Phoenician alphabet were newly made from hieratic characters, instead of being stylized versions of the original pictophonograms. However, this is an opportunity to point out that the users of the protoalphabet did take note of current Egyptian symbols and the fashion changes that occurred in the hieroglyphic script. For the sun-sign, which functioned as Shi in the syllabary and Sh in the consonantary, the hieroglyph for sun (r‘ N5) was employed in the syllabary (a circle, with or without its central dot), but it is not attested in the consonantary (though it is known in the derivative cuneiform alphabet); instead we have variations of N6, depicting the sun-disc with one uraeus serpent (New Kingdom) or two serpents (N6B, found in the MK and NK periods); the double-serpent sign is obviously the prototype for the Sinai version of Sh (with the sun disc omitted).28 Yet another variant is found twice in the Wadi el-Hol text; it has the sun-disc and the serpent, but it lacks the tail that is part of the N6 form in the New Kingdom; 29 a source for this version can be found in a surprising place, on inscriptions in Sinai (85 and 87, year 4 and 5 of the reign of Amenemhet III) which relate to Khebded, the member of the Retenu royal family who has played an important part in Orly Goldwasser’s research; it occurs as a representation of the sun in the pictures on each stela, not as a hieroglyph in the text, but it matches the protoalphabetic sign perfectly.
20. Not used by any institution or state for administrative purposes
This is an argument from silence, which ignores the possibility of lost documents; civilized administration was recorded on papyrus, which only survives in Egyptian settings. But it would be fair to say that a large number of the protoalphabetic documents on stone were official, here in Sinai, under the Egyptian government. Without being able to give a coherent and comprehensive interpretation of the corpus of inscriptions, Orly Goldwasser makes generalizations like this: it allowed the peripheral sectors of society to write their names or the name of a god, or to present a short prayer. This is true as far as it goes, but in the Semitic inscriptions at the Sinai turquoise mines there are official texts as well as private statements, though all are open to public gaze.
Four are concerned with a man named Asa: 376=1 records “the sickness (dwt) of Asa”; 345=3 is the votive sphinx from the temple, with his signature on the left shoulder above the dedicatory line, “This is my offering (nqy) to Baalat”; 358=35 inside Mine M is his obituary, “Asa has done (p‘l) his work (mlkth)”; 363=16 is on his burial site, “This grave (knkn) is the resting place (nxt) of Asa”. The block statuette from the temple (346=4) bears a prayer to Baalat “for increase of pasture (mr‘t)” (for the donkeys and goats, presumably).
The official announcements relating to Mine L, for example, were inscribed in steliform (stela-shaped) panels on the outer wall of the mine. It is reasonable to suppose that these were posted at various times and relate to different expeditions to Serabit el-Khadim. Officers are mentioned in them: “the chief prefect” (rb nçbn, which is commonly and erroneously read as rb nqbn, supposedly meaning the “chief of the miners”, from nqb “pierce, bore a hole”) appears in 349=22, and on the statuette 346=4; “the prefect of the expedition” (nçb wt.= Egp wd‘ “expedition”) is on 351=23, and possibly on 350=27.
The inscriptions concerned the equipment (’nt) for the metalworking (making and remaking the copper tools for the mining operations) by the “sons of the furnace” (bn kr), and also the vessels for watering the vegetable garden. In other places I have provided a full account of the information and instructions in the inscriptions.30 However, in our attempts to decipher these enigmatic documents, we must constantly keep in mind that the only one who really knew what an ancient alphabetic inscription meant was the person who wrote it.
21. The cuneiform alphabet was a sophisticated reworking of the protoalphabet
Goldwasser accepts that the cuneiform signs of the alphabet that was used for various purposes at Ugarit (and in other places) were based on the pictorial consonantary. However, she surmises that the protoalphabet was a despised script, which was only used by caravaneers and soldiers, but now achieved respectability in new raiment. Let us not forget that the reason we have so much written material from Ugarit is that it was preserved on clay, unlike the lost royal records of Byblos and Tyre. Clay tablets are far less fragile and perishable than papyrus rolls. If the scribes of Ugarit had chosen to write their documents on papyrus, we would have at our disposal an adze with a name and a title, another adze with the same title (rb khnm, chief priest), and a cylinder seal with a personal name on it.31 Consequently, someone would be asserting that this insignificant cuneiform script was only used for writing owners’ names on their property.
Fortunately some official documents have survived on stone and copper at Byblos (Gubla) to show that the West Semitic syllabary was used there in the Bronze Age.32 Early in the Iron Age (11th century B.C.E.) Wen Amon reported that Zakar-Baal of Byblos brought out the daybooks of his forefathers and had them read out to reveal past dealings with Egypt (was the writing syllabic or consonantal?); and 500 papyrus rolls were said to be part of the payment for a load of timber (though these have now become “smooth linen mats”).33 At Megiddo the syllabic script is found on an official signet ring (“Sealed: the sceptre of Megiddo”) from the Late Bronze Age.34 At Beth-Shemesh a scribe had made himself a copy of the cuneiform alphabet,35 and an ostracon speaks (slurringly) of carousing in a wine tavern.36
It is thus clear that all three West Semitic writing systems (syllabary, consonantary, cuneiform script) were operating around 1200 B.C.E. at the end of the Bronze Age, and the syllabary and the consonantary had flourished side by side for many centuries. There was also a new alphabetic syllabary in use in Israel.
If we think in evolutionary terms, the consonantal protoalphabet was not so much an invention as a mutation of the previous syllabic system.37 By the same token, the cuneiform alphabet was a modification of this consonantary, representing its pictorial signs with clusters of wedges (as had happened in the development of the Mesopotamian cuneiform logosyllabary). Research on the signs of any one of these three systems must always take the other two into account (as seen in the discussion of S as the sun, in section 18 above).
In the creation of the West Semitic scripts, evolution is the process; simplification is the driving force; acrophony is the creative technique, an offshoot of the older rebus principle; there is room for human intervention, but the move from syllabary to consonantary was not a new start with a new invention (acrophony) by a humble artisan who was ignorant of his own culture, as Goldwasser believes. The bulk of the letters of the protoalphabet were already functioning in the syllabary; and 18 of the 22 letters in the Phoenician alphabet of the Iron Age had an ancestor in the Canaanian syllabary of the Bronze Age (the exceptions were Het, Lamed, Sadey, Zayin). The reader should pause for a moment and ponder over this striking fact. Note further that Sadey has not been detected in a syllabic text yet, and if the tied bag (V33) was used, then it will be 19 out of 22.
The consonantal aspect of the Egyptian system had long been known to educated Semites, but it could well be that in the Middle Kingdom period, when many Phoenicians (“Asiatics”) were living in Egypt and were welcomed by the rulers, the motivational influence was there to promote further simplification in their writing, and produce the most compact system the world had seen. They might not have called this unique species of script a consonantary (or a vowelless syllabary?), but they knew how to operate the device. Its “genetic code” or “genome” contained not only letters (pictophonograms, specifically consonantograms, one of which was a double helix, incidentally) but also some lingering benign viruses, namely logograms and rebograms. In this regard, it is not necessary to suppose that the Semitic scribes focused on the monoconsonantal signs as a model, since their letters would also function as biconsonantal and triconsonantal phonograms when acting as rebograms. The use of alternative signs for particular sounds is a phenomenon arising from syllabary options (ka or ki for K, mi or mu for M; note that the signs with –a were not always given preference) or hieroglyph choices (hi for H, but three joy signs [A28, A29, A32] were available and were employed).
Ultimately Orly Goldwasser’s hypothesis was a good one, because it was falsifiable. Somebody needed to try this idea, but unfortunately it has proved to be deeply flawed, groundless rather than groundbreaking, perilously conducive to flights of fancy. Of course, the possibility remains that the protoalphabet was indeed conceived at the Sinai mines, but not on a basis of ignorance and illiteracy.
Goldwasser was striving to cause a paradigm shift in this field of study, where the consensus was certainly in need of a shake-up. When the West Semitic logosyllabary generated the logoconsonantary (the protoalphabet) there was a species of paradigm shift, and now that the process and its results are perceptible (starting with Mendenhall’s evolution insight and then the realization that the signs in both systems could also function as logograms and rebograms) we have a different approach to reading protoalphabetic inscriptions, and a fresh paradigm in the grammatological sense, that is, a table of signs and sounds which can exorcise the spectre of William Foxwell Albright, who with his imperfect stone tablets misled his followers into a barren wilderness. Orly Goldwasser was likewise misguided by the faulty chart from which the Albright school would not deviate.
Nevertheless, the things Goldwasser has achieved through the promulgation and defense of her thesis are laudable.
I have personally been stimulated to go back to all the Egyptian inscriptions from Serabit and Magharah, and simultaneously look for dating criteria to apply to the protoalphabetic and syllabic inscriptions.
Goldwasser’s colourful popularising article in Biblical Archaeology Review (2010) was enriched by helpful illustrations, notably photographs of the bilingual sphinx (345, with “beloved of Hathor” in Egyptian hieroglyphs, and “beloved of Baalat” in the West Semitic script) which paradoxically would seem to undermine her belief that the Semitic workmen could not read Egyptian writing, but it is obviously from the New Kingdom not the Middle Kingdom (for Q it has a very clear 18th Dynasty form of the cord wound on a stick, that is, V25 not V24) and so it is not relevant to the question of origins; that same article showed the Serabit temple ruins with many of the monumental inscriptions still standing, and displayed a reconstruction of this sanctuary; it opened up the subject to the public, and an enormous amount of correspondence was received by the editor; scholarly response was opposed to her ideas, notably from Anson Rainey before his death, and then Christopher Rollston (her 2012 publication, which has been scrutinized here, was a reply to him).
One correspondent, namely James E. Jennings of the University of California (Los Angeles), begged to differ on the grounds of what he had been taught by “the brilliant linguist J. Ignace Gelb”: “The Canaanites did not invent the alphabet.” 38 Right, I accept that assertion, if the emphasis is placed on the word “invent” (it was a mutation rather than an invention); but his next point is not (as Mendenhall, Hoch, and myself say) that they drew the signs out of their existing syllabary, which had already employed Egyptian hieroglyphs acrophonically for a new Semitic purpose, but this: “they extracted 22 signs from the already existing 24 uniliteral signs found in Egyptian hieroglyphic writing” (and the acrophonic principle had no part in this, according to Gelb); they produced a script that can be described as “open consonantal uniliteral writing” (whereby the sign represented a consonant plus an unspecified vowel or no vowel, and this is what I was hinting at when I mentioned “a vowelless syllabary” earlier). The figure “24” for the monoconsonantal signs is slightly suspect (24 is the number of sounds but there are alternative signs for y,w,m,n,s,t), and I have already said that only eight of the “uniliteral” or “monoconsonantal” hieroglyphs found their way into the protoalphabet (in section 18 above).
The great irony, as I see it, is that Gelb was one of the decipherers of hieroglyphic Hittite (the Luwian syllabary), but he did not realize that it had been constructed acrophonically, presumably following the pattern provided by the West Semitic logosyllabary. These days Gelb’s most faithful disciple is Barry Powell, and in his book entitled Writing (2006) he has a chapter on the origins of West Semitic writing, in which he dismisses the “discredited” acrophonic principle as a factor.39 This is quite easily done (QED): by withholding Occam’s razor (entities should not be multiplied) and regarding the “undeciphered” epigraphic material (such as our Sinai and Wadi el-Hol inscriptions) as examples of other experiments in creating scripts, and thus divorcing them from the Phoenician alphabet, as evidenced around 1000 B.C.E. when there were no picture-signs. But at that time they would be saying, so to speak, “D is for Door (Dalet)”, whereas at the start it was “Door [dalt] is for D” (or ’alp [ox] is for ’ and bayt [house] is for b) that is, practising acrophony.
Ultimately, this is the great boon that Orly Goldwasser has bestowed on her readers: people have learned that the alphabet was indeed formed by means of the acrophonic principle.40
HYPOTHETICAL ALPHABET ORIGINS
Egyptian hieroglyphs (F1 etcetera)
West Semitic acrophone (’alp etcetera)
Hebrew and Greek names of letters
’A ’alp (ox, bull) Alep (Alpha)
B bayt (house) Bet (Beta)
G gaml (boomerang) Gimel (Gamma)
D dalt (door) Dalet (Delta)
H hillul (exultation, celebration) He (Epsilon)
A28 A29 A32 (cp A8)
W waw (hook, nail, peg) Waw (Upsilon)
D dayp (eyebrow)
Z ziqq (fetter) Zayin (Zeta)
H. h.açir (mansion) Het (Eta)
H xayt. (thread) Hbr Hut. (thread, cord)
T. t.abu (good) Tet (Theta)
Z. z.il (shade) Hbr çel (shade, shadow)
Y yad (hand, forearm) Yod (Iota)
K kapp (palm, hand) kippat (palm branch) Kap (Kappa)
L lamd (training device) Hbr malmad (ox-goad) Lamed (Lambda)
S39 (crook) V1 (coil of rope)?
M maym (water) mu (water) Mem (Mu)
N naHaS (snake) Nun (Nu)
I10 (cobra) I9 (viper)
S samk (fish)
S samk (spinal column) Samek (Xi)
‘ ‘ayin (eye) Ayin (Omikron)
G Ginab (grape)
P pu (mouth) Pe (Pi)
Ç S. çirar (bag) Çade (San)
Q qaw (cord, line) Qop (Qoppa)
R ra’sh (head) Resh (Rho)
Sh shamsh (sun)
T tad (breast) Shin/Sin (Sigma)
T taw (mark) Taw (Tau)
CHART OF ALPHABET EVOLUTION
1. Gardiner, 1916, “The Egyptian Origin of the Semitic Alphabet.” Journal of Egyptian Archaeology, 3: 1–16.
2. Goldwasser, Orly 2006. “Canaanites Reading Hieroglyphs,” Ägypten und Levante 16: 121-160. Note that I prefer to use a word “Canaanian” as a synonym of “Phoenician”, since the term “Canaanite” has bad connotations from its usage in the Bible; “West Semitic” is another way of referring to the languages and scripts of Canaan (Syria-Palestine). In my opinion the terms Protosinaitic and Protocanaanite are now obsolete.
3. This is Orly Goldwasser’s reply to a correspondent (Bonnie Long) who was wondering how the Egyptian hieroglyph for N became the letter M, in Biblical Archaeology Review, 36, 5 (2010) 11. It has to be asked whether the Egyptian N manifestly represents ripples of water (it could be a range of mountain peaks); knowledge of its use would be required.
4. Goldwasser 2012, 12-14 for the most succinct statement, and almost all of the propositions being considered here are taken from there; earlier 2006, 130-156, and 2011, 263-296.
5. Colless 1988, “Recent Discoveries Illuminating the Origin of the Alphabet,” Abr-Nahrain (Ancient Near Eastern Studies) 26: 30-67; Colless 1990, “The Proto-alphabetic Inscriptions of Sinai,” Abr-Nahrain (Ancient Near Eastern Studies) 28: 1-52; Colless 1991, “The Proto-alphabetic Inscriptions of Canaan,” Abr-Nahrain (Ancient Near Eastern Studies) 29: 18-26; Colless 2010. “Proto-alphabetic Inscriptions from the Wadi Arabah,” Antiguo Oriente 8: 75-96.
6. Note that I use the informal term “protoalphabet” or “proto-alphabet” to describe the West Semitic prototype of the alphabet, which was an acrophonic consonantary, with no vowels represented, as in Egyptian writing, and like the Egyptian system it was a logo-consonantary (it had logograms) or morpho-consonantary (it had what I call rebograms, signs used as rebuses), in my understanding of it. Two useful manuals containing information on the early alphabetic inscriptions are: Sass, Benjamin 1988, The Genesis of the Alphabet and its Development in the Second Millennium B.C.; Hamilton, Gordon H. 2006. The Origins of the West Semitic Alphabet in Egyptian Scripts. With regard to the numbering system used for the Sinai inscriptions: the Egyptian texts are numbered from 1 onwards; the Sinai Semitic inscriptions are included in the same collection, beginning with Sinai 345. My additional numbering (1-44, 372=19 for example) allows easier reference to my drawings (Colless 1990, 8-11) and descriptions (1990, 12-47), and so an endnote does not need to be provided every time a Sinai inscription is mentioned. Also, Sass and Hamilton arrange their illustrations and descriptions of the inscriptions from 345 to 375d, and any particular item can readily be located in their handbooks.
7. Goldwasser 2006,153-156. Compare and contrast my table (attached to this article) with drawings of signs, and proposed hieroglyphic prototypes.
8. Sinai presumed to be the locus of the invention of the alphabet, because of the plethora of inscriptions: Goldwasser 2006, 132-133. See also Goldwasser 2012, 17 (Protosinaitic inscriptions at the mines), and 2012, 21, notes 70-73. For a larger number than 30, see Colless 1990, 8-11 (drawings of the 44 items), and 51 (table of the texts and their provenance).
9. Hamilton 2006, 289-311. Notice that Hamilton is accepting (rightly, in my view) that the inscriptions date from both the MK and the NK, but there has long been an either-or debate; see Sass 1988, 135-144.
10. Colless 2010, 91 and 95, Fig 4; http://cryptcracker.blogspot.co.nz/2009/12/wadi-el-hol-proto-alphabetic.html.
11. For an examination of these three inscriptions (from southern Egypt) comprising two “abgadaries” (inventories of the protoalphabetic letters) together with a criticism of Hamilton’s proposed identifications of the protoalphabetic signs, go to Colless, cryptcracker, Alphabet and Hieroglyphs: http://cryptcracker.blogspot.co.nz/2007/10/gordon-hamiltons-early-alphabet-thesis.html
12. Protoalphabetic inscriptions from Canaan: Colless 1991, 19-20 (listed), 22-24 (illustrated); Puech 1986, 172-187; Sass 1988, 51-75; Lemaire 2000, 110-114; Hamilton 2006, 390-400.
13. On the antiquity of the West Semitic syllabary, its use of acrophony, and its influence on the formation of the protoalphabet, see: Mendenhall, George E. 1985, The Syllabic Inscriptions from Byblos; Hoch, J. E. 1990, “The Byblos syllabary: Bridging the gap between Egyptian hieroglyphs and Semitic alphabets,” Journal of the Society for the Study of Egyptian Antiquities 20:115–124; Colless, Brian E. 1992, “The Byblos Syllabary and the Proto-alphabet,” Abr-Nahrain (Ancient Near Eastern Studies) 30: 15-62; Colless, Brian E. 1998, “The Canaanite Syllabary,” Abr-Nahrain (Ancient Near Eastern Studies) 35: 28-46; and Colless, “The West Semitic logo-syllabary” at: https://sites.google.com/site/collesseum/westsemiticsyllabary.
14. On acrophony as modified rebus writing, see Colless, Brian E. 1996. “The Egyptian and Mesopotamian Contributions to the Origins of the Alphabet,” in Guy Bunnens (ed.), Cultural Interaction in The Ancient Near East, Abr-Nahrain Supplement Series 5: 67-76; and “The evolution of the alphabet”: https://sites.google.com/site/collesseum/alphabetevolution
15. Drawings of the Gubla syllabic texts, showing the direction of writing: Colless 1993, 4 for D; 1994, 60 for C; 1994, 73 for A; 1997, 42 for G.
16. For Sinai 526 (a syllabic inscription) see Colless 1997, 47-48.
17. Valley of Queens ostracon: Goldwasser 2011, 308, Fig 3a.; Sass 1988, Fig, 286.
18. On the bow-sign as a figment, and the confusion it has caused: Colless 2010, 92, and n.48.
19. On Semitic forms of hieroglyph O6 (mansion) in Sinai 28: Goldwasser 2006, 126, Fig 6, and 144, Fig 22.
20. Photograph and drawing of Sinai inscription 53: Sass 1988, Figures 291 and 292.
21. On the fish sign as D (erroneous opinion) alongside the door sign: Goldwasser 2006, 135-137; Hamilton 2006, 61-75; Sass 1988,113-114.
22. A connection between the syllabary and the consonantary was suggested some time ago, but few have dared to explore it: Mendenhall 1985, 23-25 (“From Syllabary to Alphabet”); Colless 1992, 96-99 (“the relation of the proto-alphabet to the Byblian signary”); Colless 1998, 34-35 (comparative table of syllabic, protoalphabetic, and hieroglyphic signs).
23. Colless 2010, 82-83, 88-89, 91.
24. Colless, https://sites.google.com/site/collesseum/abgadary.
25. Colless 2010, 91 and 95, Fig 4; http://cryptcracker.blogspot.co.nz/2009/12/wadi-el-hol-proto-alphabetic.html.
26. Goldwasser 2006, 146-150.
27. Goldwasser 2011, 273-274, “The Unnecessary Hypothesis of Hieratic Sources”.
28. This point was kindly clarified to me by Stefan Wimmer (2010, 5, where he recognizes that I had shown the sun-sign connection in Colless 1988, 50-51); N6B with the two serpents was also clearly the prototype for the character on the Timna inscription, which we both studied (Wimmer 2010; Colless 2010). Incidentally, in a West Semitic logo-syllabic inscription from Thebes (New Kingdom) there is a case of the sun syllabogram (which was normally a simple circle) with a combination of serpent and disc, instead of one or the other (Colless 1997, 48-50; 1998, 31-33).
29. Wimmer 2010, 5, accepts that this is a sun-sign, and rejects the proposed connection with a composite bow; but Goldwasser (2006, 142, No 19) follows Hamilton (2006, 241-244) in this supposition that *tann (a double bow?) is the source of the letter Shin; it is true that the sign for t became Shin, but it was the human breast, tad (Colless 2010, 90 and 92).
30. For my first attempt at a comprehensive analysis of the Sinai protoalphabetic inscriptions, see Colless 1990; more recently I have placed several articles concerning the main texts on the cryptcracker website: http://cryptcracker.blogspot.co.nz/
31. See Goldwasser 2011, 292-293, “A Call from the Center: The Case of the Ugaritic Alphabet”. This cuneiform consonantary (with three syllabograms: ’a, ’i, ’u) is represented beyond Ugarit (sometimes in a reduced form, a “short alphabet”): Dietrich and Loretz 1988; Puech 1996. For the three brief inscriptions not on clay tablets, see Gordon 1965, 257 (inventory) and 159 (transcriptions). For my demonstration of the origin of the cuneiform alphabet in the signs of the protoalphabet, go to: https://sites.google.com/site/collesseum/cuneiformalphabet
32. Mendenhall 1985, 32-143; Colless 1993, 1994, 1997.
33. Breasted 1905, 106-107 (“journals” and papyrus rolls); Lichtheim 1978, 226-227 (“daybooks” and “smooth linen mats”).
34. Megiddo gold signet ring: Colless 1997, 45-46; 1998, 33; the word “sceptre” is a logogram (with hieroglyph S44 as its prototype).
35. Benjamin Sass 1991, “The Beth Shemesh Tablet,” Ugarit Forschungen 23, 315-326, provides a good introduction to this cuneiform abgadary and its ramifications; also Puech 1986, 197-213; Dietrich und Loretz 1988, 277-296.
36. Beth Shemesh ostracon: Colless 1990, 46-49; https://sites.google.com/site/collesseum/winewhine.
37. In this regard, Mendenhall 1985, 23, speaks of “the evolution from syllabary to alphabet”.
38. James E. Jennings, BAR July/August 2010, 10-12.
39. Barry B. Powell, Writing: Theory and History of the Technology of Civilization (Chichester: Wiley-Blackwell, 2006) 153-186, on West Semitic writing.
40. It needs to be added that doubt still remains whether the Egyptians knew the acrophonic principle through their single-consonant signs. Goldwasser (2012, 19, n. 1) says that these particular signs represented monosyllabic words and “did not acquire their phonetic value, as far as we know, by the use of an acrophonic procedure”. However, Jacques Freu (2000, 98), after surveying the development of ancient hieroglyphic writing, concludes that the Egyptians actually invented the alphabet, since the monoliteral signs were the precursors of the Phoenician alphabet. Freu maintains (2000, 94-95) that there was an Egyptian consonantal alphabet from the beginning, and it was constructed by the application of the principle of acrophony: for example the f of the viper sign was derived from the word ft meaning ‘viper’, and d from the cobra, dt; the water sign supplied n from nt, water; nine of the signs fit this model, and five are not clear (m, g, w, k, s); the remaining ten would fit the pattern of being monosyllabic words. It is difficult to deny that acrophony was at work here, and this principle could have been noticed by the practitioners of West Semitic writing in the Bronze Age. At the same time, this does not necessarily nullify my idea that in the evolution of West Semitic scripts, the acrophonic syllabogram and then the acrophonic consonantogram were extensions (or reductions) of the rebus principle; and these Canaanian signs continued to function as full rebograms and logograms; in this regard, Freu (2000, 95) reminds us that most of the single-sound hieroglyphs kept their ideographic value, and this could be another connection between the Egyptian and Semitic systems. Incidentally, in my estimation, only one of the protoalphabetic signs goes with a monosyllabic word, namely P (pu, mouth). Finally, it is important to remember that Charles Lenormant, in 1838, thought and taught that Phoenicians had borrowed from Egyptians the alphabetic principle and the acrophonic method (méthode acrologique), taking a selection of hieroglyphs and applying new sound values to them; and he gave as examples, the ox for ’alep, the house for B, and the eye for ‘ayin (Lemaire 2000, 105-106).
Breasted, James Henry 1905. “The Report of Wenamon,” The American Journal of Semitic Languages and Literatures 21, 2: 100-109.
Colless, Brian E. 1988. “Recent Discoveries Illuminating the Origin of the Alphabet,” Abr-Nahrain (Ancient Near Eastern Studies) 26: 30-67.
Colless, Brian E. 1990. “The Proto-alphabetic Inscriptions of Sinai,” Abr-Nahrain (Ancient Near Eastern Studies) 28: 1-52.
Colless, Brian E. 1991. “The Proto-alphabetic Inscriptions of Canaan,” Abr-Nahrain (Ancient Near Eastern Studies) 29: 18-26.
Colless, Brian E. 1992. “The Byblos Syllabary and the Proto-alphabet,” Abr-Nahrain (Ancient Near Eastern Studies) 30: 15-62.
Colless, Brian E. 1993. “The Syllabic Inscriptions of Byblos: Text D,” Abr-Nahrain (Ancient Near Eastern Studies) 31: 1-35
Colless, Brian E. 1994. “The Syllabic Inscriptions of Byblos: Texts C and A,” Abr-Nahrain (Ancient Near Eastern Studies) 32: 59-79.
Colless, Brian E. 1996. “The Egyptian and Mesopotamian Contributions to the Origins of the Alphabet,” in Guy Bunnens (ed.), Cultural Interaction in The Ancient Near East, Abr-Nahrain Supplement Series 5: 67-76.
Colless, Brian E. 1997. “The Syllabic Inscriptions of Byblos: Miscellaneous Texts,” Abr-Nahrain (Ancient Near Eastern Studies) 34: 42-57.
Colless, Brian E. 1998. “The Canaanite Syllabary,” Abr-Nahrain (Ancient Near Eastern Studies) 35: 28-46.
Colless, Brian E. 2010. “Proto-alphabetic Inscriptions from the Wadi Arabah,” Antiguo Oriente 8: 75-96.
Dietrich, Manfried, and Loretz, Oswald 1988. Die Keilalphabete: Die phönizisch-kanaanäischen und altarabischen Alphabete in Ugarit (Münster: UGARIT-Verlag).
Freu, Jacques 2000. “Les signes monolitères égyptiens, précurseurs de l’alphabet,” in Viers 2000, 87-102.
Gardiner, Alan Henderson 1916. “The Egyptian Origin of the Semitic Alphabet.” Journal of Egyptian Archaeology, 3: 1–16.
Goldwasser, Orly 2006. “Canaanites Reading Hieroglyphs,” Ägypten und Levante 16: 121-160.
Goldwasser, Orly 2010. “How the Alphabet Was Born from Hieroglyphs.” Biblical Archaeology Review 36 (2): 40-53.
Goldwasser, Orly 2011. “The Advantage of Cultural Periphery: The Invention of the Alphabet In Sinai (Circa 1840 B.C.E),” in Rakefet Sela-Sheffy and Gideon Toury (eds) Culture Contacts and the Making of Cultures (Tel Aviv: Tel Aviv University), 255-321.
Goldwasser, Orly 2012. “The Miners Who Invented the Alphabet: A Response to Christopher Rollston,” Journal of Ancient Egyptian Interconnections 4 (3): 9-22, http://jaei.library.arizona.edu
Gordon, Cyrus H. 1965. Ugaritic Textbook (Rome, Pontificium Institutum Biblicum).
Hamilton, Gordon H. 2006. The Origins of the West Semitic Alphabet in Egyptian Scripts (Washington, D.C.: Catholic Biblical Association of America).
Hoch, J. E. 1990. “The Byblos syllabary: Bridging the gap between Egyptian hieroglyphs and Semitic alphabets,” Journal of the Society for the Study of Egyptian Antiquities 20:115–124.
Lemaire, André 2000, “Les ‘Hyksos’ et les débuts de l’écriture alphabétique au Proche-Orient,” in Viers 2000, 103-133.
Lichtheim, Miriam 1978. Ancient Egyptian Literature, Volume II: The New Kingdom (Berkeley: University of California Press).
Mendenhall, George E. 1985. The Syllabic Inscriptions from Byblos (Beirut: American University of Beirut).
Puech, Émile 1986. “Origine de l’alphabet: documents en alphabet linéaire et cunéiforme du IIe millenaire,” Revue Biblique 93 (2): 161-213.
Sass, Benjamin 1988. The Genesis of the Alphabet and its Development in the Second Millennium B.C. (Wiesbaden: Harrassowitz).
Sass, Benjamin 1991, “The Beth Shemesh Tablet,” Ugarit Forschungen 23, 315-326.
Viers, Rina (ed.) 2000, Des signes pictographiques à l’alphabet. La communication écrite en Méditerranée. Actes du Colloque, 14 et 15 mai 1996 (Paris: Éditions Karthala et Association Alphabets).
Wimmer, S. J. 2010. “A Proto-Sinaitic Inscription in Timna/Israel: New Evidence on the Emergence of the Alphabet”, Journal of Ancient Egyptian Interconnections Vol. 2: 2 http://jaei.library.arizona.edu