Semitic Alphabets

Last updated:
17-Feb-2022 at 19:13
Bible chronology main page
(See History.)
Rick Aschmann
© Richard P. Aschmann
(biblechronology.net/SemiticAlphabets.pdf)

Table of Contents
1 Semitic Alphabets in North Semitic Alphabetical Order ..........................................................................................................................................................................1
1.1 Proto-Semitic...................................................................................................................................................................................................................................... 4
1.2 Proto-Sinaitic Alphabet ...................................................................................................................................................................................................................... 4
1.3 Fonts................................................................................................................................................................................................................................................... 5
1.4 Reordering of Letters .......................................................................................................................................................................................................................... 5
1.5 Arabic ................................................................................................................................................................................................................................................. 6
2 The Outcome of the Three Fricatives /š/ [ʃ], /ṯ/ [θ], and /ś/ [ɬ] in the Various Alphabets and Languages .................................................................................................6
2.1 The History of Some Semitic Fricatives ............................................................................................................................................................................................. 8
3 Hebrew Sounds Retained in the Spoken Language but not Adequately Represented in the Phoenician/Hebrew Alphabet .....................................................................9
4 Aramaic Sounds Retained in the Spoken Language but not Adequately Represented in the Alphabet................................................................................................... 10
5 A. G. Lundin’s reconstruction of “Linear Ugaritic” ............................................................................................................................................................................... 12
6 Semitic Alphabets in South Semitic Alphabetical Order ........................................................................................................................................................................ 14

Most Semitic alphabets are actually abjads, and only include consonants, not vowels. These alphabets all seem to have come from a single source (see commentary
below) except for Akkadian, which used a cuneiform syllabary rather than an abjad. (In the charts below I only list its consonants, not its actual writing system.) It is an
East Semitic language: the East Semitic languages underwent more early changes than any of the other languages, and were never written with an alphabet or abjad, but
rather with a cuneiform syllabary that was totally unrelated to any of these alphabets.

1 Semitic Alphabets in North Semitic Alphabetical Order

In the chart below, the columns containing the Semitic alphabets that followed the North Semitic alphabetical order are marked at the top of the column with ⇓.
Two South Semitic alphabets are also included for comparison; these are marked with ↓. The Ugaritic alphabet has been found in both North Semitic and South Semitic
order, and so is marked with ⇓↓. Letters in the same row generally have the same derivation. (Ugaritic letters are given in two different fonts, Aegean and MPH 2B
Damase. The former seems to be the standard form, as shown here. I am not sure why the latter has a few very divergent forms.)
At the top of each list is shown the approximate date the language or alphabet was first written, or in the case of unwritten languages, when it was probably first
spoken. This is followed by the number of consonants in each alphabet or language.
Colors used (see commentary after the chart for details):
Red: Letters which have been moved out of their standard alphabetical order in order to show their correspondence with the Ugaritic alphabet.
Pink: Consonant sounds which were not written with a distinct letter in a particular language, even though they were clearly distinguished in the spoken
language.
Yellow: Consonant sounds which were lost in a particular language, showing what other consonant they merged with.
Green: Proto-Sinaitic consonant names and shapes which were changed in the Phoenician/Hebrew alphabet.

Semitic Alphabets Last update: February 17, 2022 1 Rick Aschmann

 ↓ ↓ ⇓↓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓
Proto- IPA Ak- Old Ge’ez Name Ugaritic Trans- IPA Name Proto- Proto- An- An- Proto- An- An- Phoe- Hebrew IPA Name Mean- Early Trans- Later Trans- Arabic Trans- IPA
1
Sem- ka- Yem- Alpha- liter- Sinaitic Sinaitic cient cient Sinaitic cient cient nician Alpha- ing Aramaic liter- Aramaic liter- Alpha- liter-
itic dian eni bet ation (Lun- (Al- Name Mean- (Brian Name Mean- (& Pa- bet Alpha- ation Alpha- ation bet ation
alpha- din) bright, ing Colless) ing leo-He- bet bet
bet etc.) brew)
3750 2800 1300 100 1400 1850 1850 1850 1400 1000 300
B.C.? B.C.? B.C.? A.D. B.C.? B.C.? B.C.? B.C.? B.C.? B.C.? B.C.?
24 27 22 22
29 19 29 29 26? 27? 22 25 29 22 24? 28
(26) (28) (23) (23)

1. *ʾ ʔ – 17. �� 𐩱 13. አ ʾälf 1. 𐎀,𐎀 ʾa ʔa alpa 1. , 1. A a ʾalp ox-head 1. A a ʾalp ox 1. 𐤀 1. ‫א‬ 1. ʔ ʾaleph ox 1. ‫א‬ 1. ʾ 1. ‫א‬ 1. ʾ 1. ‫أ‬ ʾ ʔ

2. *b b 1. b 9. � 𐩨 9. በ bet 2. 𐎁,𐎁 b b beta 2. 2. B b bêt house 2. B b bayt house 2. �

� 2. ‫ב‬ 2. b beth house 2. ‫ב‬ 2. b 2. ‫ב‬ 2. ḇ/b 2. ‫ب‬ b b
throw- 3. P boom- throw-
3. *g ɡ 2. g 20. ��𐩴 20. ገ gäml 3. 𐎂,𐎂 g g gamla 3. 3. G gaml p gaml 3. �
� 3. ‫ג‬ 3. ɡ gimel 3. ‫ג‬ 3. g 3. ‫ג‬ 3. ḡ/g 3. ‫ج‬ j ɡ
stick G erang stick
hank
4. *ḫ χ 3. ḫ 14. � 𐩭 11. ኀ ḫarm 4. 𐎃,𐎃 ḫ x ẖa 4. 4. X x ḫa
of yarn
4. X x ḫayt thread 8. ‫ח‬ 4. χ 8. ‫ח‬ 4. ḫ 8. ‫ח‬ 4. ḫ 27. ‫خ‬ kh χ

5. *d d 4. d 5. D d digg fish

21. �� 𐩵 19. ደ dänt 5. 𐎄,𐎄 d d delta 5. 5. i dalt door 4. �

� 4. ‫ד‬ 5. d daleth door 4. ‫ד‬ 5. d 4. ‫ד‬ 5. ḏ/d 4. ‫د‬ d d
man
haw / calling / jubi-
6. *h h – 1. � �� 1. ሀ hoy 6. 6. H h hillul jubi-
6. H h hll
late
lation
6. 𐎅,𐎅 h h ho 5. 𐤄 5. ‫ה‬ 6. h he window 5. ‫ה‬ 6. h 5. ‫ה‬ 6. h 5. ‫ه‬ h h
w
7. *w w 5. w 6. � 𐩥 15. ወ wäwe 7. 𐎆,𐎆 w w wo 7. , 7. W w waw hook 7. W waw hook 6. 𐤅 6. ‫ו‬ 7. w waw hook 6. ‫ו‬ 7. w 6. ‫ו‬ 7. w 6. ‫و‬ w w
l
ḏiqq man- eye-
8. *z z/ 6. z (15. C c (15. C c ḏayp
→ziqq acle) brow)
weap-
dz 24. � 𐩸 17. ዘ zäy 8. 𐎇,𐎇 z z zeta 8. 8. ? ? weapon 8. � ziq fetter 7. 𐤆 7. ‫ז‬ 8. z zayin
on 7. ‫ז‬ 8. z 7. ‫ז‬ 8. z 7. ‫ز‬ z z
man- wall,
9. *ḥ ħ – 3. � 𐩢 3. ሐ ḥäwt 9. 𐎈,𐎈 ḥ ħ ḥota 9. , 9. I i ḥê(t) court 9. 𐤇 ḥasir
sion 8. 𐤇 8. ‫ח‬ 9. ħ heth
fence 8. ‫ח‬ 9. ḥ 8. ‫ח‬ 9. ḥ 8. ‫ح‬ ḥ ħ
10. (see
10. *ṭ tʼ 7. ṭ � 𐩷 21. ጠ ṭäyt 10. 𐎉,𐎉
23. � ṭ tˤ ṭet 10. 10. ? ṭô(t) spindle?
Colless)
ṭab good 9. 𐤈 9. ‫ט‬ 10. tˤ ṭeth wheel 9. ‫ט‬ 10. ṭ 9. ‫ט‬ 10. ṭ 9. ‫ط‬ ṭ tˤ
11. (see
11. *y j 8. y � 𐩺 18. የ yämän 11. 𐎊,𐎊
26. � y j yod 11. , 11. Y y yad arm
Colless)
yad hand 10. �
� 10. ‫י‬ 11. j yodh hand 10. ‫י‬ 11. y 10. ‫י‬ 11. y 10. ‫ي‬ y j
palm
12. *k k 9. k 12. �𐩫 14. ከ kaf 12. 𐎋,𐎋 k k kaf 12. , 12. K k kapp palm 12. K k kap palm 11. 𐤊 11. ‫כ‬ 12. k kaph (of 11. ‫כ‬ 12. k 11. ‫כ‬ 12. ḵ/k 11. ‫ك‬ k k
hand)
(see row
13. 𐎌,𐎌 š ʃ šin 13. 26. *š below)
13. L l 13. L
13. *l l 10. l 2. �� 𐩡 2. ለ läwe 14. 𐎍,𐎍 l l lamda 14. lamd ox-goad lamd goad 12. 𐤋 12. ‫ל‬ 13. l lamedh goad 12. ‫ל‬ 13. l 12. ‫ל‬ 13. l 12. ‫ل‬ l l
J j J j
14. *m m 11. m 4. �� 𐩣 4. መ may 15. 𐎎,𐎎 m m mem 15. 14. M m mêm water 14. M m maym water 13. �
� 13. ‫ מ‬14. m mem water 13. ‫מ‬ 14. m 13. ‫מ‬ 14. m 13. ‫م‬ m m
ḏiqq eye-
15. *ḏ ð (6. z) 25. �
�𐩹 17. ዘ 16. 𐎏,𐎏 ḏ ð ḏal 16. , 15. C c manacle 15. C c ḏayp (7. ‫)ז‬ 6. ‫ז‬ 15. ḏ (8. ‫)ח‬ d 26. ‫ذ‬ dh ð
→ziqq brow
n n ser-
16. *n n 12. n 13. ��𐩬 12. ነ nähas 17. 16. N naḥš snake 16. N naḥš snake 14. 𐤍 14. ‫נ‬ 15. n 14. ‫נ‬ 16. n 14. ‫נ‬ 15. n 14. ‫ن‬ n n
f f pent
17. 𐎐,𐎐 n n nun nun fish

θʼ/ 17. (see

17. *ṱ (15. ṣ) 29. �� 𐩼 23. ጸ 18. 𐎑,𐎑 ẓ ðˤ ẓu 18. (18. E e ʿên eye) ẓil shade (18. ‫)צ‬ 18. ‫צ‬ 17. ṱ (9. ‫)ט‬ (ṭ ) 24. ‫ظ‬ ẓ zˤ
tθʼ Colless)

Semitic Alphabets Last update: February 17, 2022 2 Rick Aschmann

 ↓ ↓ ⇓↓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓
Proto- IPA Ak- Old Ge’ez Name Ugaritic Trans- IPA Name Proto- Proto- An- An- Proto- An- An- Phoe- Hebrew IPA Name Mean- Early Trans- Later Trans- Arabic Trans- IPA
1
Sem- ka- Yem- Alpha- liter- Sinaitic Sinaitic cient cient Sinaitic cient cient nician Alpha- ing Aramaic liter- Aramaic liter- Alpha- liter-
itic dian eni bet ation (Lun- (Al- Name Mean- (Brian Name Mean- (& Pa- bet Alpha- ation Alpha- ation bet ation
alpha- din) bright, ing Colless) ing leo-He- bet bet
bet etc.) brew)

18. *s s 13. s 17. ? samk fish 18. D d samk fish

(see sup-
11. �𐩪 7. ሰ sat 19. 𐎒,𐎒 s s samka 19. Colless)
samk
port 15. 𐤎 15. ‫ס‬ 16. s samekh support 15. ‫ס‬ 18. s 15. ‫ס‬ 16. s 15. ‫س‬ s s

19. *ʿ ʕ – 18. �
�𐩲 16. ዐ ʿäyn 20. 𐎓,𐎓 ʿ ʕ ʿain 20. 18. E e ʿên eye 19. E ʿên eye 16. 𐤏 16. ‫ע‬ 17. ʕ ʿayin eye 16. ‫ע‬ 19. ʿ 16. ‫ע‬ 17. ʿ 16. ‫ع‬ ʿ ʕ
20. *p p 14. p 21. 19. P p piʾt corner?

16. �� 𐩰 25. ፈ äf 21. 𐎔,𐎔 p p pu 20. e pu mouth 17. �

� 17. ‫פ‬ 18. p pe mouth 17. ‫פ‬ 20. p 17. ‫פ‬ 18. p̄ /p 17. ‫ف‬ f f
sʼ/
21. *ṣ 15. ṣ 28. � 𐩮 23. ጸ ṣädäy 22. 𐎕,𐎕 ṣ sˤ ṣade 22. , 20. S s ṣa(d) plant 21. Q q ṣirar bag 18. �
� 18. ‫צ‬ 19. sˤ tsadhe hunt 18. ‫צ‬ 21. ṣ 18. ‫צ‬ 19. ṣ 18. ‫ص‬ ṣ sˤ
tsʼ
ɬʼ/
22. *ṣ́ 19. �� 𐩳 24. ፀ ṣ́äppä ↑ (29. ?)�� ↑ ↑ ↑ 19. ‫ק‬ 22. ṣ́ (16. ‫)ע‬ (ġ ) 25. ‫ض‬ ḍ dˤ
tɬʼ
22. (see cord, needle
23. *ḳ kʼ 16. q 5. �
� 𐩤 8. ቀ ḳaf 23. 𐎖,𐎖 q q qopa 23. 21. Q q qu(p) monkey
Colless)
qaw
line 19. 𐤒 19. ‫ק‬ 20. q qoph
head 19. ‫ק‬ 23. q 19. ‫ק‬ 20. q 19. ‫ق‬ q q

24. *r ɾ 17. r 8. �
� 𐩧 6. ረ rəʾs 24. 𐎗,𐎗 r r raša 24. , 22. R r raʾš head 23. R r raʾš head 20. 𐤓 20. ‫ר‬ 21. r resh head 20. ‫ר‬ 24. r 20. ‫ר‬ 21. r 20. ‫ر‬ r ɾ
com-
25. *ṯ θ2 ↓ 27. �� 𐩻 7. ሰ s 25. 𐎘,𐎘 ṯ θ ṯanna 25. 23. t ṯann posite 24. t ṯad breast ↓ ↓ 21. ‫ש‬ 25. ṯ (22. ‫)ת‬ (t ) 28. ‫ث‬ th θ
bow
25. (see
26. *š ʃ2 18. š 15. �� 𐩯 7. ሰ s 24. V v šimš sun
Colless)
šimš sun (15. ‫)س‬ s s
13. 𐎌,𐎌 š ʃ šin 13. 21. �
� 21. ‫ש‬ 22. ʃ shin tooth 21. ‫ש‬ 26. š 21. ‫ש‬ 22. š
com-
27. *ś ɬ2 ↑ 7. �� 𐩦 5. ሠ śäwt (30. 𐎝,𐎝 3 ś) (s ) śu (28. ) (24. t ṯann posite (25. t ṯad Breast) ↑ 21. ‫ש‬ 23. ɬ sin 21. ‫[ש‬ɬ] 27. ś (15. ‫)ס‬ (s ) 21. ‫ش‬ sh ʃ
bow)

28. *ġ ʁ – 22. �𐩶 16. ዐ ʿ 26. 𐎙,𐎙 ġ ɣ ġain 26. , 25. F ġa ? 26. ? ġinab grape (16. 𐤏) 16. ‫ע‬ 24. ʁ 16. ‫ע‬ 28. ġ (16. ‫)ע‬ 23. ġ 23. ‫غ‬ gh ʁ
owner’s owner’s
29. *t t 19. t 10. �
�𐩩 10. ተ täwe 27. 𐎚,𐎚 t t to 27. 26. T taw
mark
27. T taw
mark 22. 𐤕 22. ‫ת‬ 25. t taw mark 22. ‫ת‬ 29. t 22. ‫ת‬ 24. ṯ/t 22. ‫ت‬ t t

28. 𐎛,𐎛 ʾi ʔi i

29. 𐎜,𐎜 ʾu ʔu u
(see row
30. 𐎝,𐎝 s2(ś) śu 28. 27. *ś above)

𐎟,𐎟 word divider

1
In Wiktionary names are provided for most of these letters, though meanings and Ugaritic spellings are not provided. However, it turns out that these names are merely reconstructions,
mainly from the later Phoenician/Hebrew names, or in a few cases are simply the syllable pronunciations, at least according to this page. Even so, apparently these reconstructions are based
on good evidence that Ugaritic really did have names for their letters, and that we know for certain at least the first syllable of most of the letters, based on a tablet called KTU 5.14, which
contains most of the letters with a corresponding Akkadian equivalent showing this first syllable. This tablet and the conclusions drawn from it are shown on this page.
2
See The Outcome of the Three Fricatives /š/ [ʃ], /ṯ/ [θ], and /ś/ [ɬ] to understand the ins and outs of these sounds and the letters used to represent them.
3
This letter was apparently not used in Ugaritic to represent a separate sound, but based on its appearance in the South Semitic alphabetical order, it was evidently intended to write the [ɬ]
sound in other Semitic languages, corresponding to the sound traditionally transcribed as /ś/ in Hebrew and in Proto-Semitic.

Semitic Alphabets Last update: February 17, 2022 3 Rick Aschmann

 The dates in the large chart above are all estimates, and should all be taken with a big grain of salt. For the most part they are not my estimates, but estimates of
experts in their respective fields. Even so, every one of them is a guess. For the languages that were written, this is usually the estimated date of the first texts found in ar-
cheological digs, but this does not mean that these languages might not have been written earlier, it simply indicates that neither archeology nor secular history can give
evidence that it was written earlier. For Proto-Semitic the date is even more of a guess, because only comparative linguistics gives us any information, and neither archeol-
ogy nor history can tell us anything about it.
I have put the date for the Phoenician alphabet down as 1400 B.C., which is much earlier than most sources state, and have cited an article about the 22-letter
Cuneiform Short Alphabet (archived link) as evidence. This alphabet has an identical consonant inventory as the Phoenician alphabet, clearly distinguishing it from the 27-
letter Ugaritic alphabet, suggesting that it was used to write a Semitic language with only 22 consonants, probably Phoenician.
Of the 29 consonants in Proto-Semitic, Arabic retained 28 (but see Arabic below) and Ugaritic 27, though if we include 30. 𐎝/𐎝 for /ś/ [ɬ] there are letters for 28,
though not the same 28 as for Arabic.
1.1 Proto-Semitic
Proto-Semitic was not a written language, and in fact evidently predated the first writing by many hundreds of years, but based on comparison of all the known
languages descended from it, it seems clear that it had a total of 29 consonants, and the Old Yemeni or South Semitic Alphabet (and its variant the Old North Arabian
alphabet shown in the South Semitic Alphabetical Order section) retained all 29 of them (though with a couple of significant changes in pronunciation, shown in the South
Semitic chart in red in the IPA column). See Reflexes of Proto-Semitic Sounds in Daughter Languages and the link for the first column in the chart. All of these alphabets,
including the South Semitic ones, were probably derived from the same alphabet, which must be at least as old as the Ugaritic alphabet, which shows both alphabetical
orders.
1.2 Proto-Sinaitic Alphabet
As many have suggested, the Proto-Sinaitic alphabet is the most likely candidate for this original alphabet, and both the North Semitic and South Semitic alphabets
were evidently derived from it, including Ugaritic. The information for Proto-Sinaitic in the chart was obtained from various sources. The Wikipedia article suggests that
there is doubt about Proto-Sinaitic being an alphabet, but actually there is little doubt, as shown by this article and the Colless article mentioned below.
However, there seem to be (at least) two drastically different analyses of the inscriptions, and these affect the final alphabet inventory and the letter names. I have
presented both in the chart: the first is by William Albright and others, and the other is by Brian Colless.
The Albright, etc. analysis is listed first, and uses a Proto-Sinaitic font obtained from ancientroadpublications.com/Fonts.html#ProtoSinaitic, which was based on
Albright’s Schematic Table of Proto-Sinaitic Characters found at the bottom of this page. The Ancient Names and Meanings mostly follow this table and the modified
table by Simon Ager on this page. Some additional ideas can be found on this page (in Spanish), but I did not find them all that helpful.
The Colless analysis can be found beside this in the chart, and is based on this article, and especially on the chart at the bottom of it, which was drawn by hand and
is somewhat hard to read. For many letters I have been able to use the same font as for Albright’s analysis, though some of the symbols have been reinterpreted, but some-
times I could not, in which case I have said “(see Colless)”, meaning that you will need to consult Colless’s chart. (I have not yet had a chance to read Colless’s article ex-
haustively, and I very much want to, because it sounds like he has done his homework.) I mention Colless repeatedly here: search to see all comments.
Proto-Sinaitic was evidently used to write an earlier form of a South Canaanite dialect, with more consonants than later Hebrew, Canaanite, or Phoenician. How
many letters (consonants) did Proto-Sinaitic have? There is no way to know, since the data available is very limited. However, if we assume for each of these two analyses
that the analysis is accurate, and if we also include letters that clearly existed later on in Phoenician and Hebrew (several are missing for Albright, and only one for Col-
less) we get a list of 26 letters for Albright and 27 for Colless, as shown in the chart above. Actually, Albright, Colless, and Ugaritic line up quite well: the one Proto-
Semitic consonant totally missing from Ugaritic (22. */ṣ́/ [ɬʼ/tɬʼ]) is also missing from both Albright, etc. and Colless’s lists, and Ugaritic 25. */ṯ/ [θ] apparently corresponds
to at least South Canaanite /ś/ [ɬ] (see footnote 2 above). This strongly suggests that the 27 letters of the basic Ugaritic system also represent the complete Proto-Sinaitic
alphabet.
Of course, there could have been more consonants written in Proto-Sinaitic that have simply not come to light because of the limited number of inscriptions, and it
is even possible that all of the letters in the Old Yemeni or South Semitic Alphabet came directly from Proto-Sinaitic, in which case it would have had symbols for all of
the 29 consonants of Proto-Semitic. This is not outside the realm of possibility, since languages in both the far north (Early Aramaic) and in the far south (Old South Ara-

Semitic Alphabets Last update: February 17, 2022 4 Rick Aschmann

bian) did retain all of these consonants, and Canaan appears to be the area of greatest innovation involving loss of consonants, but when those losses occurred is impossible
to determine. (This page assumes that Proto-Sinaitic had 30 letters, but they are evidently equating its alphabet with that of Ugaritic, which did have 30 letters, though it
only had 27 native consonants.)
Even if it is true that the Proto-Sinaitic alphabet had all 29 letters, we have no way of knowing where the two letters missing from the 27-letter Ugaritic alphabet
would have been placed in the North Semitic Alphabetical order, or even if they were ever placed in such an ordering. These are Proto-Semitic 27. /ś/ [ɬ], which probably
corresponds to letter 30 in the Ugaritic alphabet, but was clearly an afterthought because it was not a sound used in Ugaritic (having merged with /s/ [s]), and Proto-Semitic
22. /ṣ́/ [ɬʼ/tɬʼ], which had merged with 21. /ṣ/ [sʼ/tsʼ] and so was not listed in the Ugaritic alphabet.
However, one fact which suggests that there was a form of the Proto-Sinaitic alphabet that had all 29 letters is precisely the fact that the Old Yemeni or South Se-
mitic Alphabet did have all 29 consonants, and in the next section I suggest that we might use this alphabet as the Proto-Alphabet of Proto-Sinaitic.
One factor entering into this question is that some of the letters may have changed their names from Proto-Sinaitic to Phoenician/Hebrew (at least in the case of the
Albright, etc. analysis): the Phoenician alphabet page says, «according to a theory by Theodor Nöldeke from 1904, some of the letter names were changed in Phoenician
from the Proto-Canaanite script.» (Apparently the letter shape was usually also different as well.) The list follows, with some comments by me. I have colored these letters
green (under the Albright, etc. analysis), and I have attempted to determine, simply by looking at the letter shapes, which of the two options the various alphabet symbols
and letter names seem to correspond to, putting the two options on separate lines separated by a grey line. These are quite subjective, but perhaps instructive!

gaml “throwing stick” → gimel “camel” I’m skeptical about this one, that it ever meant camel, as is Wikipedia: it looks like all they did was flip the symbol, so
I left them all on one line!
digg “fish” → dalet “door”
hll “jubilation” → he “window”
ziqq “manacle” → zayin “weapon” Apparently ziqq was originally ḏiqq, and was the name of the letter pronounced /ḏ/, which later merged with /z/ in
Canaanite.
naḥš “snake” → nun “fish” According to this page, only the name was changed, not the symbol.
piʾt “corner” → pe “mouth”
šimš “sun” → šin “tooth” See The Outcome of the Three Fricatives /š/ [ʃ], /ṯ/ [θ], and /ś/ [ɬ] below.

I have not yet studied Colless’s article and data sufficiently to do the same for it, but I have colored his letters green when they disagree with Albright, etc., and
have tried to arrange them in this system as well.
1.3 Fonts
The Aegean font used for Ugaritic was found on this page. The MPH 2B Damase font used for Ugaritic and Phoenician was found on this page. The two fonts
used for the Old Yemeni or South Semitic Alphabet (Sabaic and Qatabanic styles) were found on this page. However, since I have now posted this file in PDF format, all
fonts should appear correctly for all users without having to download them.
1.4 Reordering of Letters
Arabic letters in red are normally at the end of the alphabet, as the numbering shows, but I have arranged them to show their relation to Ugaritic, since these two
languages retained more of the original Semitic consonants than did most Canaanite languages (including Phoenician and Hebrew). These letters were placed at the end
because the Arabic Alphabet was derived from the Aramaic Alphabet (essentially identical to the Hebrew Alphabet), but since this alphabet did not have all of the sounds
in Arabic, these six letters were invented just for Arabic.
Because of the complex way in which phonemes have merged, I have also moved two of the Ugaritic phonemes out of their place in the alphabetical order, and
have marked them in red also. (See Footnote 2 above for an explanation of the red letter in the Proto-Sinaitic column.)

Semitic Alphabets Last update: February 17, 2022 5 Rick Aschmann

 1.5 Arabic
Arabic only lost one of the 29 Proto-Semitic consonants, 26. */š/ [ʃ], which merged with 18. */s/, but it changed the pronunciation of a number of them, which
confuses the comparison with the other Semitic languages, as shown in the chart below. In this chart I only include significant sound changes, not minor differences like
ejective versus pharyngeal or uvular.

Proto-Semitic Sounds which were Retained in Arabic, but with significant pronunciation change.
Proto- IPA Arabic Trans- IPA
Sem- Alpha- liter-
itic bet ation
3. *g ɡ 3. ‫ج‬ j ɡ This is pronounced [dʒ] in most Arabic varieties, but is still [ɡ] in Egyptian Arabic.
17. *ṱ θʼ/ tθʼ 24. ‫ظ‬ ẓ zˤ
20. *p p 17. ‫ف‬ f f
22. *ṣ́ ɬʼ/ tɬʼ 25. ‫ض‬ ḍ dˤ
27. *ś ɬ 21. ‫ش‬ sh ʃ
Thus all cases of 26. */š/ [ʃ] became a simple [s] in Arabic, merging with existing [s], but then all cases of 27. */ś/ [ɬ] changed to [ʃ], which corresponded to a sound
in Hebrew which later came out as [s], which is terribly confusing!

2 The Outcome of the Three Fricatives /š/ [ʃ], /ṯ/ [θ], and /ś/ [ɬ]
in the Various Alphabets and Languages
Based on its shape and pronunciation, the Phoenician letter 21. � (pronounced /š/ [ʃ]), from which came the Hebrew letter ‫ש‬, used for two Hebrew consonants ‫ש‬
(/š/ [ʃ]) and ‫( ש‬/ś/ [ɬ]), was apparently derived from Ugaritic 13. 𐎌/𐎌 (/š/ [ʃ] šin), but it is placed alphabetically in the place of Ugaritic 25. 𐎘/𐎘 (/ṯ/ [θ] ṯanna). (I don’t
know why the two different fonts show such incredibly distinct forms for Ugaritic 25; the former seems to be the right one, and neither looks anything like the Phoenician
or Hebrew letter.) However, both Phoenician 21. � (/š/ [ʃ]) and Ugaritic 13. 𐎌/𐎌 (/š/ [ʃ] šin) look a lot like the Proto-Sinaitic letter t, which was evidently pronounced
as either /ś/ [ɬ] or as /ṯ/ [θ], not as /š/ [ʃ], and both Albright and Colless list a different letter for /š/ [ʃ]. So how do we make sense of this?
Now the Old Yemeni or South Semitic Alphabet, and its variant the Old North Arabian alphabet, contain symbols for all of these sounds, in fact for all of the
original Proto-Semitic consonants, so the simplest solution might be to assume that the South-Semitic alphabet represents the Proto-Alphabet from which all of the others
came. The problem is that in the South-Semitic alphabet the symbol 7. 𐩦 � is used to represent neither /š/ [ʃ] nor /ṯ/ [θ], but /ś/ [ɬ], precisely the sound that was lost in both
South Canaanite and Ugaritic!
However, it is precisely the variations in the two alphabetical orders, North Semitic and South Semitic, that suggest an explanation and confirm that the South Se-
mitic alphabet is probably the original. In the South Semitic Alphabet letter 7. 𐩦 � represents /ś/ [ɬ] and letter 15. 𐩯 � represents /š/ [ʃ], but in the South Semitic
alphabetical order established for the Ugaritic alphabet this order is reversed: letter 7. 𐎌/𐎌 represents /š/ [ʃ] and letter 15. 𐎝/𐎝 was apparently intended to represent /ś/ [ɬ]
in other languages, even though Ugaritic had lost this sound. And as we saw above, the Phoenician alphabet also shows a change in the alphabetical order for the
Phoenician letter 21. � /š/ [ʃ]. To me this suggests that the Old Yemeni or South Semitic Alphabet shows us the original use of the letters it contains, but that the symbol �
or �, originally used for the sound /ś/ [ɬ], was later reassigned to a different sound in Ugaritic, which no longer had the /ś/ [ɬ] sound, and replaced the original symbol for
/š/ [ʃ].
In Phoenician this new symbol 21. � for /š/ [ʃ] ended up being the only symbol that was retained, since /ṯ/ [θ] and /ś/ [ɬ] had both fallen together with /š/ [ʃ].
However, Hebrew (and North Canaanite in general) did not lose /ś/ [ɬ], but because Phoenician ended up being the parent alphabet for Hebrew, this symbol 21. � had to do
double duty for both /š/ [ʃ] and /ś/ [ɬ]. (See Hebrew Sounds Retained in the Spoken Language.)

Semitic Alphabets Last update: February 17, 2022 6 Rick Aschmann

 In fact, Phoenician turns out to be the only Canaanite language that merged all three sounds. But it turns out that South Canaanite and Ugaritic on the one hand and
North Canaanite on the other ended up slightly different as regards the distribution of these three sounds. In Hebrew the Semitic phoneme */ṯ/ [θ] became /š/ [ʃ] (later
spelled ‫)ש‬, which was a distinct sound from /ś/ [ɬ] (later spelled ‫)ש‬. However, this page says that «according to William Albright, [the letter sin or shin] was based on a
“bow” and with the phonemic value /š/ [ʃ] “corresponds etymologically (in part, at least) to original Semitic /ṯ/ (th), which was pronounced /s/ [sic] in South Canaanite”»
(Presumably he means /ś/ [ɬ] here, later pronounced s in Hebrew.) This seems to suggest that */ṯ/ [θ] merged with /ś/ [ɬ] in South Canaanite but with /š/ [ʃ] in Central
Canaanite and Hebrew. If so, then this explains why Albright’s table mentioned below shows the letter ‫( ש‬/ś/ [ɬ]) as derived from a Proto-Sinaitic letter t with the name
/ṯann/ (apparently equivalent to Ugaritic ṯanna “composite bow”), but shows a different Proto-Sinaitic letter V /šimš/ “sun” (later Hebrew ‫ ֶׁש ֶׁמש‬/ˈšemeš/) for the /š/ [ʃ]
sound. In Phoenician all three of the sounds /ṯ/ [θ], /š/ [ʃ], and /ś/ [ɬ] were merged into one and were spelled with the Phoenician letter �, so the letter V was not retained
in Phoenician or the later Aramaic and Hebrew alphabets which derived from it, even though the sounds /š/ [ʃ] and /ś/ [ɬ] were still distinguished in Hebrew and Aramaic
and were later spelled ‫ ש‬and ‫ש‬. As for the symbol t, it looks like both a bow (or, according to Colless, like breasts) and like a couple of teeth, so was apparently renamed
/shin/ by reidentification.
Confused? No wonder!

Perhaps the table on the next page entitled “The History of Some Semitic Fricatives” can help clarify the situation.
(In the table, unless specified otherwise, I obtained the Proto-Semitic and Hebrew forms of possible letter names from ahdictionary.com/word/semitic.html and
from shodhganga.inflibnet.ac.in:8080/jspui/bitstream/10603/37097/2/appendices.pdf. The connection of ‫“ ָׁשנָׁ ה‬bend” with */ṯann/ “composite bow” is mostly guesswork,
but the second link does confirm that this verb had */ṯ/ [θ] in Proto-Semitic.)
Brian Colless has a different opinion about the meaning, names, and origin of the symbols used, but otherwise his analysis agrees with the arrangement in the ta-
ble.

The Proto-Alphabet Letter 15. 𐩯 � and its Unique History

This letter for the /š/ [ʃ] sound was named “sun” and given its symbol accordingly, but according to Wiktionary and this note, the original Proto-Semitic form of the word “sun”
was */śamš/, not */šamš/, with a different initial consonant, but the initial /ś/ [ɬ] became /š/ [ʃ] in the Northwest Semitic languages (including Aramaic and all Canaanite languages,
including Ugaritic). The second link clarifies that this change occurred because of “regressive assimilation” from the following /š/ [ʃ], which was a process that only occurred in Northwest
Semitic, and which occurred even in cases like this where the consonants are not adjacent. What makes the Wiktionary entry confusing is that every single form is shown with initial /š/ [ʃ]
except the Proto-Semitic form, which initially made me doubt its validity, but in fact in every case /š/ [ʃ] is the expected outcome for /ś/ [ɬ], and in some cases would not have been the
outcome if the initial consonant had been /š/ [ʃ]. (See the first chart above and the last chart below to verify this.) For example, in Arabic */š/ [ʃ]→/s/ [s], but */ś/ [ɬ]→/š/ [ʃ], so Proto */š/ [ʃ]
never corresponds to Arabic /š/ [ʃ]. Even more confusing, In Old South Arabian */š/ [ʃ]→[s̪ ], an unusual retracted sibilant which was spelled with ��, leaving these languages without the /š/
[ʃ] sound, and so /ś/ [ɬ] is often written /š/ in transcribing words in these languages, even though it was still pronounced [ɬ].
In any case, this word was evidently pronounced /šimš/ in Proto-Sinaitic, and lent its name and letter shape to the Proto-Sinaitic letter V and to the corresponding South Semitic
letter 15. 𐩯� �. Further evidence for this is that the Old North Arabian variant of � is , which looks exactly like a sun! The Old South Arabian and Old North Arabian languages adopted the
entire alphabet, including this letter to write the /š/ [ʃ] sound (and its later reflex [s̪ ] in Old South Arabian). They kept the sun symbol for this sound, even though their word for sun did not
begin with this sound, because they didn’t invent it, only adopted it. (In fact the word for “sun” in Old South Arabian was ���, according to the Wiktionary entry, whose consonants would
have been /śms/ [ɬms]. I don’t know why the last consonant changed from � to �.)
It is impossible to know what name the users of the South Semitic alphabet used for this letter, because the letter names are only known from Ge’ez, not from any earlier South
Semitic languages, but Ge’ez had lost this sound and so discarded the letter!
(ahdictionary.com/word/semitic.html and shodhganga.inflibnet.ac.in/bitstream/10603/37097/2/appendices.pdf do not agree that the original Proto-Semitic stem began with /ś/ [ɬ],
the former claiming that the consonantal stem was */šmš/, and the latter that it was */šms/ or */šmš/, but neither seems to have considered the implications of the Arabic and Old South
Arabian forms as I have explained above.)

Semitic Alphabets Last update: February 17, 2022 7 Rick Aschmann

 A. G. Lundin has proposed a different “Proto-Alphabet”, which often follows the South Semitic one, but not always, and definitely not in the case of these three
fricatives. (See the last column in the table on the next page and the discussion on Lundin later on.) In particular, he theorizes that South Semitic 15. 𐩯 � corresponds to
Ugaritic 30. 𐎝/𐎝 and represents a fish, but all we need to do is look at the North Arabian descendant 16. of South Semitic 15. 𐩯 � to see that it clearly was a sun and not a
fish! (See the following box for more info.)
2.1 The History of Some Semitic Fricatives

Proto- Proto-Semitic Proto- Old Proto-Sinaitic Later Ugaritic Phoenician North Lun-
Semitic Form of Alpha- North and South Sym- (Far-North Canaanite din’s
Probable Later bet (& Ara- Canaanite bol Canaanite) & Hebrew “Proto-
Letter Name South bian Use Alpha-
Semit- Sym- bet”
ic)a bol
(Only one symbol
26. */š/ [ʃ] *śamš “sun”→šamš b 15. 𐩯� � 16. V šimš “sun”  (symbol lost) ‫ ֶׁש ֶׁמש‬/ˈšemeš/
retained)
13. 𐎌/𐎌 /š/ [ʃ] šin  21. � /š/ [ʃ] shin ‫ ש‬/š/ [ʃ] shin
*šinn “tooth”   7. 𐩦� � 13.
“tooth” “tooth”   “tooth” (‫ ֵׁ֖שן‬šên)
*ṯann (sounds � or t: ṯann 25. 𐎘/𐎘 /ṯ/ [θ] ṯanna  (‫ ָׁשנָׁ ה‬/šāˈnāh/
25. */ṯ/ [θ] 27. 𐩻� � 28.  27. 𐩻� � 25.
“composite bow” merged “composite bow” “composite bow” “to bend”???)
*ṯad “female in South or ṯad “female ‫ ַׁשד‬šaḏ “female
 
breast(s)” Canaanite breast(s)” breast(s)”
Original name and Uga- (Pronunciation (� symbol
unknown, possibly ritic, both uncertain, eitherreassigned
śäwt (Ge’ez alphabet), symbols /ṯ/ [θ] or /ś/ [ɬ])
to a different
27. */ś/ [ɬ] 7. 𐩦� � 7.   ‫ ש‬/ś/ [ɬ] śin [ˈɬin]
or first syllable possibly retained sound with
śu1 (Ugaritic alphabet), for a time?) a new name)
meaning unknown  
in either case. (new symbol created for Ugaritic)  (30. 𐎝/𐎝) 28.

Note that none of these sounds was the same as the simple [s] sound, which was quite distinct and experienced no changes:
‫ ס‬/s/ [s] samekh
21. 𐤎 /s/ [s] samekh
18. */s/ [s] ? 11. 𐩪� � 11. 𐩪� � 19. 𐎒/𐎒 /s/ [s] samka (‫ ָׁס ַׁמך‬sāˈmaḵ a)
“support”
“to lean, support”
However, much later, well after the completion of the Old Testament, Hebrew ‫ ש‬/ś/ [ɬ] changed to [s], giving ‫ ש‬and ‫ ס‬the same pronunciation.

a
I show the South Semitic symbols here, because this is the only alphabet that retained the full inventory, but this alphabet was developed for a Northwest-Semitic language, probably
Proto-Sinaitic, not for a South Semitic language. The importance of this is seen in the next footnote.
b
See the box above entitled The Proto-Alphabet Letter 15. 𐩯 � and its Unique History for an explanation of this letter’s name.

Semitic Alphabets Last update: February 17, 2022 8 Rick Aschmann

 3 Hebrew Sounds Retained in the Spoken Language but not
Adequately Represented in the Phoenician/Hebrew Alphabet
In earlier versions of this article I had said that Hebrew had retained only 23 of the 29 Proto-Semitic consonants, since the distinction between [ʃ] and [ɬ] was
clearly retained and eventually distinguished in the spelling, as ‫ ש‬and ‫ ש‬even though the alphabet officially has only 22 letters. Thus ‫ ש‬and ‫ ש‬are normally listed as one
letter in the alphabet, but have different pronunciations and dotting.c Eventually the pronunciation of ‫ ש‬merged with ‫ס‬, but the spelling distinction was retained.d The
original pronunciation of ‫ ש‬was probably [ɬ], which is how Sylvester the cat pronounces his s’s in the Warner Brothers cartoons.e This was determined by comparing He-
brew words containing ‫ ש‬with the same words in some of the South Semitic languages that retain this sound. However, there are also a few vestiges of evidence in Greek
words borrowed from Hebrew:
Early Later Examples
Proto-Semitic Greek
Hebrew Hebrew Early Hebrew f
Akkadian Greek meaning
*/ś/ [ɬ] ‫ ש‬/ś/ [ɬ] ‫ ש‬/ś/ [s] [ls] λσ ‫ ֵׁ֖ב ֶׁשם‬/ˈbōśem/ [ˈboːɬem] βάλσαμον [ˈbálsamɔn] balsam
*/ś/ [ɬ] ‫ ש‬/ś/ [ɬ] ‫ ש‬/ś/ [s] [l] λ ‫ ַׁכ ְׂש ִּֽדים‬/kaśˈdîm/ [kaɬˈdîm] kaldu Χαλδαῖοι [xalˈdâiɔi] Chaldeans

The Hebrew word ‫ ֵׁ֖ב ֶׁשם‬/ˈbōśem/ [ˈboːɬem] is quite frequent in the Old Testament, usually translated “spice” in modern translations, but the Greek word βάλσαμον
is not used to translate it in either the Septuagint or the Greek New Testament.
The Greek word Χαλδαῖοι derives from the Akkadian, in which Proto-Semitic *[ɬ] is normally realized as /š/ [ʃ], but according to William Barrick: “Akkadian
scholars have long recognized a peculiarity of the Akkadian language: the phenomenon of a phonetic shift of the sibilant (‫ ש‬/ś/ [ɬ]) to a lamed when the sibilant is followed
by a dental (‫ ד‬/d/).” This is not surprising given the original pronunciation, but would make little sense with a pronunciation of /š/ [ʃ].

However, it seems that Hebrew actually retained 25 consonant phonemes, as seen in the IPA column to the right of the Hebrew Alphabet Column in the large chart
above, not just 23, using ‫ ח‬and ‫ ע‬to write two consonant phonemes each. The reason they only wrote with 22 letters is because they adopted the Phoenician alphabet to
write Hebrew, and the Phoenician dialect of Canaanite had already reduced its consonant inventory to 22. The following chart is taken from
en.wikipedia.org/wiki/Biblical_Hebrew_language#Consonants, but I have modified it to make it easier to read:

c
This dotting was not used to distinguish these two sounds until the second half of the first millennium A.D., so the distinction was only maintained orally up to that time, just as it was for
the two pronunciations of ‫ ח‬and ‫ע‬. The difference is that the latter merged into one sound each before the dotting was invented, whereas the pronunciation distinction of ‫ ש‬and ‫ ש‬is still
maintained today.
d
It might seem that the story of “shibboleth” ‫ ש ֹּ֜בלֶת‬/šibˈbōleṯ/ and “sibboleth” ‫ ס ֹּ֗בלֶת‬/sibˈbōleṯ/ in Judges 12:6 provides evidence that these sounds were already merging at the time of the
judges, and that the merger had progressed farther in some regions than in others. However, this is an apparent case of the merger of ‫ ש‬/š/ [ʃ] and ‫ ס‬/s/ [s], not of ‫ ש‬/ś/ [ɬ] and ‫ ס‬/s/ [s],
which is confusing, since there is no evidence that ‫ ש‬/š/ [ʃ] and ‫ ס‬/s/ [s] had merged in any Northwest Semitic language of this period, according to Hendel (jstor.org/stable/1357296), and
they are still clearly distinguished in Hebrew today. However, Hendel’s article provides a proposed solution, which turns out to have no bearing on our discussion here.
e
In this example the tongue is not pulled back as much as it is for [ɬ] in most of the languages that use it, but it is still clearly [ɬ] rather than [s].
f
In Hebrew and Aramaic the text between slashes is the traditional transliteration, which matches the written Hebrew form with a late (Tiberian) pronunciation, whereas the text between / /
is the probable real pronunciation in IPA. I have not always been able to complete this information for Aramaic.

Semitic Alphabets Last update: February 17, 2022 9 Rick Aschmann

 Later Examples
Proto-Semitic Aramaic Arabic
Hebrew Hebrew f
Early pron. Late pron. Aramaic f Arabic meaning
‫ חמ ָׁשה‬/ḥămišˈšâ/ [χamiʃˈʃaːh] [χamiʃˈʃaːh] ‫ ַׁח ְׂמ ָׁשא‬/ḥaməˈšāʾ/ [χaməˈʃaːʔ] ‫ˈ[خمسة‬xamsah] five
*/ḫ/ [χ] [χ] ‫خ‬
‫ ח‬/ḥ/ [ħ] ‫ ח‬/ḥ/ [ħ] ‫ ָׁצ ַׁרח‬/ṣāˈraḥ/ [sˤaːˈraχ] [sˤaːˈraħ] ‫צרח‬ ‫[صرخ‬sˤaˈrax] shout (verb)
*/ḥ/ [ħ] [ħ] ‫ح‬ ‫ ָׁמלַׁ ח‬/māˈlaḥ/ [maːˈlaħ] [maːˈlaħ] ‫מלח‬ ‫ˈ[ملح‬milħ] salt (verb)
‫ ערב‬/ʿōˈrêḇ/ [ʁoːˈreb] [ʕoːˈreb] ‫עראב‬ ‫[غراب‬ʁaˈraːb] raven
*/ġ/ [ʁ] [ʁ] ‫غ‬
‫ ע‬/ʿ/ [ʕ] ‫ ע‬/ʿ/ [ʕ] ‫ ַׁמע ָׁרב‬/maʿăˈrāḇ/ [maʁaˈraːb] [maʕaˈraːb] ‫ערב‬ ‫ˈ[غرب‬ʁarb] west
*/ʿ/ [ʕ] [ʕ] ‫ع‬ ‫ ֶׁע ֶׁבד‬/ˈʿeḇeḏ/ [ˈʕebed] [ˈʕebed] ‫ עבד‬/ˈʿăḇêḏ/ [ˈʕabēd] ‫ˈ[عبد‬ʕabd] slave

“…these phonemes are also distinguished consistently in the Septuagint of the Pentateuch…, but this becomes more sporadic in later books and is generally absent
in Ezra and Nehemiah.”
Numerous examples can also be found in the Greek New Testament:
Later Examples
Proto-Semitic Arabic Greek
Hebrew Hebrew f
Early pron. Late pron. Arabic g Greek meaning
*/ḫ/ [χ] [χ] ‫خ‬ [x] χ ‫ ָׁרחל‬/rāˈḥêl/ [raːˈχeːl] [raːˈħeːl] (‫[ )راحیل‬raːˈħiːl]) Ῥαχήλ [r̥axéːl] Rachel
‫ ח‬/ḥ/ [ħ]
*/ḥ/ [ħ] [ħ] ‫ح‬ — ‫ י ְׂצ ָׁחק‬/yiṣˈḥāq/ [yisˤˈħaːq] [yisˤˈħaːq] ‫[ إسحق‬ʔisˈħaːq] Ἰσαάκ [isaák] Isaac
‫ ַׁעזָׁ ה‬/ʿazˈzâ/` [ʁazˈzaːh] [ʕazˈzaːh] ‫ˈ[ غزة‬ɣazzah] Γάζα [ɡáza] Gaza
*/ġ/ [ʁ] [ʁ] ‫غ‬ [ɡ] γ ‫מֹורה‬
ָׁ ‫ ע‬/ʿămōˈrâ/ [ʁamoːˈraːh] [ʕamoːˈraːh] (‫[ عمورة‬ʕamuːˈrah]) Γόμοῤῥα [ɡɔ́mɔr̥a] Gomorrah
‫ ע‬/ʿ/ [ʕ] Χοδολλογομόρ
‫ ְׂכ ָׁד ְׂרלָׁ ֹ֙ע ֶׁמ ֹ֙ר‬/kədārəlāˈʿōmer /[kədaːrəlaːˈʁoːmer] /[kədaːrəlaːˈʕoːmer] Chedorlaomer
[xɔdollɔɡɔmɔ́ɾ]
*/ʿ/ [ʕ] [ʕ] ‫ع‬ — ‫ ע ָׁשו‬/ʿēˈśāw/ [ʕēˈɬaːw] [ʕēˈɬaːw] ‫[ عِیسُو‬ʕiˈsu] Ἠσαῦ [eːsâu] Esau

Thus these four phonemes were still distinguished at the time of the writing of the Septuagint, in the 3rd century B.C., after the completion of the Old Testament,
and it seems probable that the [ɬ] pronunciation of ‫ ש‬was retained to around that time also. However, by the time the Masoretes developed their diacritic system for clarify-
ing the pronunciation of Tiberian Hebrew in the second half of the first millennium A.D., these distinctions had been lost, and these sounds had their Later Hebrew
pronunciation.

4 Aramaic Sounds Retained in the Spoken Language

but not Adequately Represented in the Alphabet
The situation of Aramaic is even more complicated. Its alphabet was borrowed in about the year 1000 from the Phoenician Alphabet, which only had 22 letters,
and as stated in Note 4 under en.wikipedia.org/wiki/Semitic_languages#Consonants:

g
Those examples in parenthesis in Arabic do not have the expected letter, but since these are names, they were probably borrowed from Hebrew at a late stage.

Semitic Alphabets Last update: February 17, 2022 10 Rick Aschmann

 Although early Aramaic (pre-7th century BCE) had only 22 consonants in its alphabet, it apparently distinguished all of the original 29 Proto-Semitic pho-
nemes, including */ḏ/, */ṯ/, */ṱ/, */ś/, */ṣ́/, */ġ/ and */ḫ/ — although by Middle Aramaic times, these had all merged with other sounds. This conclusion is
mainly based on the shifting representation of words etymologically containing these sounds; in early Aramaic writing, the first five are merged with /z/,
/š/, /ṣ/, /š/, /q/, respectively, but later with /d/, /t/, /ṭ/, /s/, /ʿ/. (Also note that due to begadkefat spirantization, which occurred after this merger, OAm.
/t/→/ṯ/ and /d/→/ḏ/ in some positions, so that PS */t/,/ṯ/ and */d/,/ḏ/ may be realized as either of /t/,/ṯ/ and /d/,/ḏ/ respectively.) The sounds */ġ/ and */ḫ/
were always represented using the pharyngeal letters ʿ and /ḥ/, but they are distinguished from the pharyngeals in the Demotic-script papyrus Amherst 63,
written about 200 BC. This suggests that these sounds, too, were distinguished in the Old Aramaic language, but written using the same letters as they later
merged with.

Thus of all the alphabets listed here, Aramaic’s was the one least suited to the language it represented, with 29 consonants in Early Aramaic and 24 in Later Ara-
maic being represented by only 22 letters. (Like Hebrew there was a distinction between the sounds later written as ‫ ש‬and ‫ ש‬in the Aramaic sections of the Old Testament,
but as in Hebrew these dots were not added until the second half of the first millennium A.D.) Those extra consonant sounds that were not distinguished are marked in pink
on the chart.

Semitic Alphabets Last update: February 17, 2022 11 Rick Aschmann

 5 A. G. Lundin’s reconstruction of “Linear Ugaritic”
A. G. Lundin (archived link) suggests that the source of the Ugaritic cuneiform alphabet was not cuneiform but a linear alphabet that had a form very close (though
not identical) to the “Proto-Alphabet” of all the Semitic languages, presumably derived from Proto-Sinaitic, though he does not make this entirely clear. This “Linear
Ugaritic” corresponded exactly to the Ugaritic alphabet, and in fact he claims to be able to reconstruct the actual forms of the “Linear Ugaritic” letters based on the
Ugaritic wedge shapes (with help from the later alphabets). His conclusions (Table I, page 94) are shown in the chart below, with the numbering corresponding to that of
the Ugaritic Alphabet in the chart above, with 27 consonants plus 3 extras added later. His system is quite logical and compelling, though of course that does not make it
authoritative! However, Brian Colless (different article than the one mentioned repeatedly above) seems to agree with many of his conclusions, suggesting that he should
be taken seriously.
Lundin’s theory, then, is that Ugaritic was originally written in a linear alphabet like the other Semitic Languages, but that because it was written on perishable
materials no examples have been found. This alphabet was then converted to a cuneiform alphabet in order to write it on clay, a much more permanent medium.
Cuneiform had only two signs: 1. The wedge, usually 𐎚 or 𐎂, though it could have other orientations
2. The angle wedge 𐎓 (Winkelhaken), in other fonts 𐎓.

The following are Lundin’s rules for how the Linear Ugaritic letters were transformed into cuneiform, sometimes rephrased by me to make them clearer. The
numbers represent the consonant numbers in the chart. To really understand how the rules work, you need to read the article.

Rule 1: A circle in Linear Ugaritic is represented by the angle wedge in Ugaritic cuneiform (10, 18, 20, 23, 25; Figure 1).
Rule 2: A straight line (or a line with an appendix) in Linear Ugaritic is represented by a wedge in Ugaritic cuneiform. Vertical and horizontal lines retain their ori-
entation, but oblique lines are shown in various ways.
Rule 3: A broken or curved line in Linear Ugaritic is represented by three successive wedges in Ugaritic cuneiform (17, 4, 11). (However, sometimes three succes-
sive wedges represent a straight line of three segments as in 5.) Figure 3 shows both cases.
Rule 4: Two parallel wedges in Ugaritic cuneiform usually represent an angle in Linear Ugaritic, not parallel lines (21, 18, 19, 12, 24; Figure 4, misnumbered as
Figure 5).
Rule 5: Crossed lines in Linear Ugaritic are represented by two wedges meeting at a right angle (10, 15?; Figure 5, misnumbered as Figure 4). The rest of this rule
seems confusing to me, or based on only one real example.

These rules are unidirectional: the cuneiform letters can be derived from the linear letters, but not the reverse. Thus Lundin had to use all of the known linear al-
phabets as well as the cuneiform to help him reconstruct the presumed original alphabets.

In the chart below, PA, Lundin’s reconstructed “Proto-Alphabet”, is essentially the same as the Proto-Sinaitic Alphabet, which others have estimated from the 17th
–13th century B.C., except that Lundin claims it had symbols for the full 29 consonants of Proto-Semitic. However, in his chart he does not list *ṣ́ [ɬʼ/tɬʼ], because this is
the one phoneme which had no symbol in the Ugaritic alphabet. Lundin recognizes this, though he calls this phoneme /ḍ/, which is actually its Arabic reflex. Presumably
its South Semitic form is the only information we have on this letter, so I have added this at the bottom of the chart.
Beside Lundin’s chart I am showing each Ugaritic consonant in two different fonts. The first font (“Aegean”) is usually closer to Lundin’s analysis, and also seems
to be the standard form; the second is the one I have originally used everywhere else in this article, though I have now included both.
Beside these I also show the corresponding South Semitic (Old Yemeni) consonant, again in two different fonts, corresponding to the Sabaic and Qatabanic varia-
tions. Lundin seems to have mostly followed the Sabaic forms. As expected, the one Proto-Semitic consonant not listed is 22. *ṣ́ [ɬʼ/tɬʼ], corresponding to South Semitic 19.
South Semitic symbol 27. 𐩻 �, which corresponds to Ugaritic 25. */ṯ/ [θ], looks quite different from Lundin’s symbol , but apparently Lundin’s was a standard
variation, since it matches the Old North Arabian symbol 28. ..
South Semitic 29. 𐩼 �, which corresponds to Ugaritic 18. */ṯ/ [θʼ], is also a bit different from Lundin’s symbol , though its correspondence looks plausible.

Semitic Alphabets Last update: February 17, 2022 12 Rick Aschmann

 A. G. Lundin’s reconstruction of “Linear Ugaritic” and the “Proto-Alphabet”

F1 F2 SSS SSQ z F1 F2 SSS SSQ

1. 𐎀 𐎀 17. �
� 𐩱 1. 𐎏𐎏
16. 25. �� 𐩹 16.

2. 𐎁 𐎁 9. � 𐩨 2.

3. 𐎂 𐎂 20. �
� 𐩴 3. 17.𐎐 𐎐 13. �� 𐩬 17.
4. 𐎃 𐎃 14. � 𐩭 4. 18.𐎑 𐎑 29. �
� 𐩼 18.
5. 𐎄 𐎄 21. �� 𐩵 5. 19.𐎒 𐎒 11. � 𐩪 19.
6. 𐎅 𐎅 1. � � 6. 20.𐎓 𐎓 18. �
� 𐩲 20.
7. 𐎆𐎆 6. � 𐩥 7. 21.𐎔 𐎔 16. �
� 𐩰 21.
8. 𐎇 𐎇 24. � 𐩸 8. 22.𐎕 𐎕 28. � 𐩮 22.

9. 𐎈 𐎈 3. � 𐩢 � 9. 23.𐎖 𐎖 5. �� 𐩤 23.
10. 𐎉 𐎉 23. �
� 𐩷 10. 24.𐎗𐎗 8. � � 𐩧 24.
11. 𐎊 𐎊 26. �� 𐩺 11. 25.𐎘 𐎘 27. �� 𐩻 25.
12. 𐎋 𐎋 12. � 𐩫 12. 26.𐎙 𐎙 22. � 𐩶 26.

13. 𐎌 𐎌 7. �� 𐩦 13.
27.𐎚 𐎚 10. �
� 𐩩 27.
14. 𐎍𐎍 2. �� 𐩡 14. 28.𐎛 𐎛
17. �
� 𐩱 (1.)
15. 𐎎 𐎎
4. �� 𐩣 15. 29.𐎜 𐎜

The dates for U.Lin. and PA are Lundin’s estimates. The

other dates are merely those of the first documented examples, and 30.𐎝 𐎝 15. �
� 𐩯 28.
in fact Lundin suspects that the 29-consonant SS variety existed in
Palestine and Syria as early as the 15th century B.C. 19. ��𐩳 29.
ṣ́ �? �?
Column Titles: Signe = Proto-Semitic consonant PS = Phoenician Script (for official or monumental use)
F1 = Font 1 (Aegean) U = Ugaritic (14th –13th century B.C.) SS = South Semitic Alphabet (10th century B.C., aka the Old Yemeni Alphabet)
F2 = Font 2 (MPH 2B Damase) U.Lin. = Lundin’s reconstructed “Linear Ugaritic” PA = Lundin’s reconstructed Proto-Alphabet (late 16th – early 15th century B.C.)
SSS = South Semitic Alphabet, Sabaic font P = Phoenician (12th –11th century B.C.) Black numbers (except parenthesized) represent the established alphabetical
SSQ = South Semitic Alphabet, Qatabanic font Parc. = Partial?? orders. The red numbers represent the 29 consonants of Proto-Semitic, PA, and SS.

Semitic Alphabets Last update: February 17, 2022 13 Rick Aschmann

 Lundin says that Ugaritic 1. */ʾ/ [ʔ], corresponding to South Semitic 17, had two different forms, both representations of a bull’s head, represented by the
Phoenician letter 𐤀, and the South Semitic one 17. 𐩱 � (though again, his form doesn’t look exactly like the ones in the fonts, though again the Old North Arabian form
does!). He theorizes that both were used in Ugaritic in order to provide a way to write the three vowels, for [ʔa], and flipped sideways to provide forms to represent [ʔi]
and [ʔu].

6 Semitic Alphabets in South Semitic Alphabetical Order

In the chart below the columns containing the actual South Semitic alphabets are marked at the top of the column with ⇓. The Ugaritic alphabet has been found in
both North Semitic and South Semitic order, and so is marked ⇓↓; it is numbered in the South Semitic order, but the North Semitic order is also shown for reference. Let-
ters in the same row generally have the same derivation. (Ugaritic letters are given in two different fonts, Aegean and MPH 2B Damase. The former seems to be the stand-
ard form, as shown here. I am not sure why the latter has a few very divergent forms.)
There is some variation in the South Semitic order, as shown in the variation in numbering. The Ugaritic order is probably the original, but because the Old Yem-
eni or South Semitic Alphabet is the only one (along with the Old North Arabian alphabet which was derived from it) with a complete inventory of the original Semitic
consonants, I have listed them in this order, marking variations in Ugaritic, Ge’ez, and Old North Arabian in red. (A couple of the South Semitic consonants had significant
changes in pronunciation from their Proto-Semitic counterparts, and these are shown in the chart below in red in the IPA column)
The Modern South Arabian languages are apparently not descended from the languages that used the Old Yemeni Alphabet, all of which died out no later than 600
A.D., though they are related to them. Even so, in both cases all of the original 29 Proto-Semitic consonants are retained, with only a couple of changes in pronunciation in
the Modern South Arabian languages (marked in red in the chart below).
For the Old North Arabian alphabet (a derivative of the Old Yemeni or South Semitic Alphabet) there are absolutely no free fonts available to represent it! How-
ever, I was able to find images for the letters. You can see the characters themselves and a description at unicode.org/charts/PDF/Unicode-7.0/U70-10A80.pdf, pages 2 and
3. This alphabet was not included in the North Semitic chart, since it does not differ essentially from the Old Yemeni or South Semitic Alphabet, though one letter is
placed differently.
The web sites unicode.org/charts/PDF/U10A60.pdf and unicode.org/charts/PDF/Unicode-7.0/U70-10A80.pdf give names for all of the letters in the Old Yemeni or
South Semitic Alphabet and in the Old North Arabian alphabet respectively, but I believe that these names are simply names of convenience taken from various actual
names used in other alphabets, or modifications of these, and do not reflect actual knowledge of their original names. I have included these in the chart below, but have
colored them yellow to show that they are not reliable.

Colors used:
Red: Letters which have been moved out of their standard alphabetical order in order to show their correspondence with the Old Yemeni or South Semitic
Alphabet.
Pink: New letters not derived from previous alphabets, representing innovative consonant sounds.
Yellow: Consonant sounds which were lost in a particular language, showing what other consonant they merged with. (Also used for letter names that are
assumed to not reflect the actual names in a particular alphabet, as explained above.)

The only languages that still use a form of the South Semitic Alphabet today are those spoken in Ethiopia, which use modifications of the Ge’ez alphabet. As an
example of these I have included the Amharic alphabet. It still retains all of the letters in the Ge’ez alphabet, even though it has lost (or merged) a number of the sounds. It
is actually a syllabary, with the vowels attached in various ways to the original Ge’ez letters, but I have listed the original forms without an attached vowel (which in the
Amharic syllabary actually represent syllables with the vowel normally transliterated as ä).

Semitic Alphabets Last update: February 17, 2022 14 Rick Aschmann

 ⇓ ⇓ ⇓ ⇓↓ ↓
Proto- IPA Modern Old trans- IPA Name Old North Name Ge’ez trans- IPA Name Amharic IPA Ugaritic trans- IPA Name North
1
Semitic South Yemeni liter- Arabian Alphabet liter- Alphabet Alphabet liter- Semitic
Arabian Alphabet ation Alphabet ation ation order
3750 ? 1300 100 100 1400
B.C.? B.C.? B.C.? A.D. B.C.?
29 29 29 29 29 24 28?
(26)

� 1. ሀ 1. ሀ h
1. *h h [h] 1. � h h he 1. heh h h hoy 1. 𐎅/𐎅 h h ho 6

2. � 2. ለ 2. ለ l
2. *l l [l] 𐩡 l l lamedh 2. lam l l läwe 2. 𐎍/𐎍 l l lamda 14

ħ [ħ] ḥ ħ 3. ሐ ḥ ħ ḥäwt 3. ሐ ḥ ħ
3. *ḥ 3. � 𐩢 heth 3. hah h 3. 𐎈/𐎈 ḥota 9

meem 4. መ 4. መ
4. *m m [m] 4. � 𐩣 m m mem 4. m m may m 4. 𐎎/𐎎 m m mem 15

kʼ [kʼ] 5. � 8. ቀ ḳ kʼ ḳaf 9. ቀ kʼ
5. *ḳ 𐩤� q q qoph 5. qaf 5. 𐎖/𐎖 q q qopa 23

6. � 15. ወ 20. ወ w 6. 𐎆/𐎆

6. *w w [w] 𐩥 w w waw 6. waw w w wäwe w w wo 7

ɬ 7. � ɬ 5. ሠ ś ɬ śäwt 5. ሠ s š ʃ
7. *ś [ɬ] 𐩦 s 2 shin 7. es-2 7. 𐎌/𐎌 4 šin 13

ɾ 8. � 6. ረ 6. ረ r
8. *r [r] 𐩧 r r resh 8. reh r r rəʾs 8. 𐎗/𐎗 r r raša 24

9. � 9. በ 10. በ b
9. *b b [b] 𐩨 b b beth 9. beh b b bet 14. 𐎁/𐎁 b b beta 2

10. � 10. ተ 12. ተ

10. *t t [t] 𐩩 t t taw 10. teh t t täwe t 9. 𐎚/𐎚 t t to 27

11. � 7. ሰ 7. ሰ s 10. 𐎒/𐎒

11. *s s [s] 𐩪 s1 s sat 11. es-1 s s sat s s samka 19

12. � 14. ከ 18. ከ k

12. *k k [k] 𐩫 k k kaph 12. kaf k k kaf 11. 𐎋/𐎋 k k kaf 12

13. � 12. ነ 15. ነ n

13. *n n [n] 𐩬 n n nun 13. noon n n nähas 12. 𐎐/𐎐 n n nun 17

χ ḫ 11. ኀ ḫ χ ḫarm 14. ኀ ḫ

14. *ḫ [x] 14. � 𐩭 x kheth 14. khah h 13. 𐎃/𐎃 x ẖa 4

15. *š ʃ [ʃ],[h] 15. � 𐩯 s3 s̪ samekh 16. es-3 7. ሰ s s (15. 𐎝/𐎝 4) ś śu 30

16. � 25. ፈ 33. ፈ

16. *p p [f] 𐩰 f f fe 17. feh f f äf f 16. 𐎔/𐎔 p p pu 21

ʔ 17. � ʾ ʔ 13. አ ʾ ʔ ʾälf 17. አ ʔ ʾ ʔa

17. *ʾ [ʔ] 𐩱 ʾalef 18. ʾalef 17. 𐎀/𐎀 alpa 1

ʕ 18. � ʿ ʕ 16. ዐ ʿ ʕ ʿäyn 21. ዐ ʔ ʿ ʕ

18. *ʿ [ʕ] 𐩲 ʿayn 19. ʿain 18. 𐎓/𐎓 ʿain 20

ɬʼ/tɬʼ 19. � ḍ ɬˤ 24. ፀ ṣ́(ḍ) ɬʼ ṣ́äppä 32. ፀ tsʼ ṣ sˤ

19. *ṣ́ [ɬʼ] 𐩳 dhadhe 20. dad (𐎕/𐎕) (22)

ɡ ɡ 20. ገ ɡ 27. ገ
20. *g [ɡ] 20. � 𐩴� g gimel 21. geem g gäml g 20. 𐎂/𐎂 g g gamla 3

21. � 19. ደ 25. ደ

21. *d d [d] 𐩵 d d daleth 22. dal d d dänt d 21. 𐎄/𐎄 d d delta 5

22. *ġ ʁ [ɣ] 22. � 𐩶 ġ ɣ ghayn 23. ghain 16. ዐ ʿ ʕ ʿäyn 22. 𐎙/𐎙 ġ,ʿ ɣ ġain 26

tʼ [tʼ] 23. � ṭ tˤ 21. ጠ ṭ tʼ ṭäyt 28. ጠ tʼ ṭ tˤ

23. *ṭ 𐩷 ṭeth 24. ṭah 23. 𐎉/𐎉 ṭet 10

24. � 17. ዘ 22. ዘ z

24. *z z/dz [z] 𐩸 z z zayn 25. zain z z zäy 24. 𐎇/𐎇 z z zeta 8

25. *ḏ ð [ð] 25. � 𐩹 ḏ ð dhaleth 26. thal 25. 𐎏/𐎏 ḏ→d ð ḏal 16

26. � 18. የ 24. የ

26. *y j [j] 𐩺 y j yodh 27. yeh y j yämän j 26. 𐎊/𐎊 y j yod 11

27. *ṯ θ [θ] 27. � 𐩻 ṯ θ thaw 28. theh 7. ሰ s s 27. 𐎘/𐎘 ṯ θ ṯanna 25

Semitic Alphabets Last update: February 17, 2022 15 Rick Aschmann

 ⇓ ⇓ ⇓ ⇓↓
Proto- IPA Modern Old trans- IPA Name Old North Name Ge’ez trans- IPA Name Amharic IPA Ugaritic trans- IPA Name North
1
Semitic South Yemeni liter- Arabian Alphabet liter- Alphabet Alphabet liter- Semitic
Arabian Alphabet ation Alphabet ation ation order
sʼ/tsʼ [sʼ] ṣ sˤ 23. ጸ ṣ tsʼ ṣädäy 31. ጸ tsʼ ṣ sˤ
28. *ṣ 28. � 𐩮 ṣadhe 15. ṣad 28. 𐎕/𐎕 ṣade 22

29. *ṱ θʼ/tθʼ [θʼ] 29. � 𐩼 ẓ θˤ theth 29. zah 19. 𐎑/𐎑?ẓ/ðˤ ṱ→ġ ðˤ ẓu 18

22. ጰ P̣ pʼ̣ pạ̈ yt 30. ጰ pʼ ʾi ʔi

𐩽 No. 1 1 (29. 𐎛/𐎛) i 28

26. ፐ 34. ፐ p ʾu ʔu
𐩾 No. 50 10 p p psa (30. 𐎜/𐎜) u 29
numeric 8. ሸ ʃ
𐩿 indicator
20 𐎟/𐎟 word divider

11. ቨ v
13. ቸ tʃ
16. ኘ ɲ
19. ኸ h(x)
23. ዠ ʒ
26. ጀ dʒ
29. ጨ tʃʼ

4
If we compare lines 7 and 15 of the South Semitic chart above, we see that the Old Yemeni sounds and the Ugaritic sounds seem to have swapped places. However, in fact Ugaritic 𐎝/𐎝,
numbered 15 in the South Semitic alphabetical order and 30 (as if an afterthought) in the North Semitic order, was apparently not used in Ugaritic to represent a separate sound, but was
either unused or was used for the same sound as 10. 𐎒/𐎒 [s]. Therefore it is not surprising that the Ugaritic alphabet swapped 7 and 15. This is why it seems probable that Ugaritic 𐎝/𐎝 was
intended to write the [ɬ] sound in other Semitic languages, traditionally transcribed /ś/. For more info, see The Outcome of the Three Fricatives /š/ [ʃ], /ṯ/ [θ], and /ś/ [ɬ].

Semitic Alphabets Last update: February 17, 2022 16 Rick Aschmann